2-Amino-4,6-diarylpyridines as novel ligands for the estrogen receptor

Article abstract of DOI:10.1016/S0960-894X(01)00321-3

We have prepared a novel series of 2-amino-4,6-diarylpyridines that function as ligands of estrogen receptor alpha (ERα) and estrogen receptor beta (ERβ). These compounds bind to both ERα and ERβ with a modest selectivity for the alpha subtype. The most p

Full text of DOI:10.1016/S0960-894X(01)00321-3

Bioorganic & Medicinal Chemistry Letters 11 (2001) 1939–1942
2-Amino-4,6-diarylpyridines as Novel Ligands for the Estrogen
Receptor
Brad R. Henke,a,* David H. Drewry,a Stacey A. Jones,c Eugene L. Stewart,b
Susan L. Weaverc and Robert W. Wiethea
aDepartment of Medicinal Chemistry, Glaxo Wellcome Research and Development, Five Moore Drive,
Research Triangle Park, NC 27709, USA
bDepartment of Structural Chemistry, Glaxo Wellcome Research and Development, Five Moore Drive,
Research Triangle Park, NC 27709, USA
cDepartment of Molecular Endocrinology, Glaxo Wellcome Research and Development, Five Moore Drive,
Research Triangle Park, NC 27709, USA
Received 29 March 2001; accepted 8 May 2001
Abstract—We have prepared a novel series of 2-amino-4,6-diarylpyridines that function as ligands of estrogen receptor alpha (ERa)
and estrogen receptor beta (ERb). These compounds bind to both ERa and ERb with a modest selectivity for the alpha subtype.
The most potent of these analogues, compound 19, has a Ki=20 nM at ERa. These molecules represent a novel template for
designing potentially useful ligands for the estrogen receptor. # 2001 Elsevier Science Ltd. All rights reserved.
Estrogen is an effective treatment for both menopausal
symptoms and the prevention and management of
postmenopausal osteoporosis. Despite the beneficial
effects of estrogen, there is evidence to suggest an
increase in reproductive tissue cancer,1 which leads to
both a restriction of widespread use and long-term
compliance issues. Thus, many pharmaceutical compa-
nies have engaged in the development of agents that can
maintain the benefits of estrogen while avoiding the
risks. The development of selective estrogen receptor
modulators (SERMs) such as raloxifene and tamoxifen
(Fig. 1) that have tissue-selective agonist or antagonist
effects are agents that hold the promise of a safer
alternative to estrogen.2,3
transcriptional properties.5 Additionally, studies with
receptor subtype-specific knockout mice suggest that the
two ER subtypes have distinct biological roles.6
A detailed picture of both the binding requirements and
the mode of action of steroidal and nonsteroidal ER
ligands has been developed through the use of molec-
ular modeling and X-ray crystallographic analysis of
agonist- and antagonist-bound ERa and ERb.7À10 The
ER binds a wide range of steroidal and nonsteroidal
ligands with moderate to high affinity, with a minimal
requirement of at least one phenol as the basic pharma-
cophore. The structural diversity of nonsteroidal estro-
gens having affinity for the ER is remarkable, with a
number of chemically distinct templates being reported
as ER ligands.2,3 As part of our ongoing effort in the
identification of novel ligands for nuclear receptors, we
Estrogens exert their biological effects via the estrogen
receptor (ER), a protein that functions as a ligand-
modulated gene transcription factor. The recent dis-
covery of a second ER subtype, termed ERb, has
increased the level of complexity of estrogen signaling.4
The two receptor subtypes (ERa and ERb) show sig-
nificant sequence homology in their DNA and ligand
binding domains; however, they exhibit differences in
their tissue distribution patterns, ligand selectivity, and
*Corresponding author. Fax: +1-919-315-0430; e-mail: brh14990@
gsk.com
Figure 1. Representative ER ligands.
0960-894X/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(01)00321-3

1940
B. R. Henke et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1939–1942
sought to develop structurally novel templates that are
readily amenable to parallel synthesis. Herein we report
the synthesis and biological activity of a novel series of
2-amino-4,6-diarylpyridine ER ligands.
achieved using methodology previously described by
Katritzky.13 Treatment of the chalcone derivatives with
an equimolar amount of 2-(benzotriazol-1-yl)acetoni-
trile and an excess (20 equiv) of the desired secondary
amine in refluxing ethanol afforded the desired 2-amino-
4,6-diarylpyridines in modest yields. It is worthwhile to
note that while this reaction worked reasonably well
with simple secondary amines, we were unable to isolate
any of the desired 2-aminopyridine cyclization products
when anilines or primary amines were employed as
nucleophiles. While the failure of anilines to engage in
this cyclization may be attributable to their poor
nucleophilic character, the lack of success with primary
amines remains puzzling. In addition, secondary amines
containing a free hydroxyl group also failed to afford
any desired product. Deprotection of the THP acetals
was achieved by treatment with 80% aqueous acetic
acid at 60 ꢀC. The entire reaction sequence was readily
The preparation of compounds 1–19 (Table 1) is out-
lined in Scheme 1. The various substituted 4-hydroxy
acetophenones were protected as their corresponding
THP acetals using sulfuric acid adsorbed on silica gel as
catalyst,11 which allowed isolation of the product in
high yield and purity by simple filtration of the catalyst
and removal of solvent. Formation of the chalcone
intermediates was best carried out by treating a 1 M
ethanol solution containing equimolar amounts of the
desired aromatic aldehyde and acetophenone with 0.25
equiv of finely ground solid NaOH.12 Other methods of
chalcone formation were less reliable and gave lower
yields. Formation of the 2-aminopyridine nucleus was
Table 1. In vitro profile of ER ligands 1–19
Compd
Structurea
R3
Bindingb
R1
R2
R4
ERa Ki (nM)
ERb Ki (nM)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
H
H
H
m-CH3
H
H
p-OH
o-CH3
H
H
H
H
H
H
H
H
H
H
H
H
H
H
H
–CH3
–CH3
–CH3
–CH3
–CH3
–CH3
–CH3
–(CH2)4–
–CH3
–CH3
–(CH2)2-Ph
–(CH2)2-Ph
–(CH2)2-Ph
490Æ10 (2)
>3000 (2)
650Æ50 (2)
160Æ10 (4)
620Æ60 (2)
210Æ50 (2)
660Æ85 (2)
IA
380Æ10 (4)
450Æ20 (2)
IA
1260Æ50 (2)
180Æ20 (4)
70Æ10 (4)
IA
830Æ140 (4)
130Æ10 (2)
2000Æ30 (2)
2 Æ10 (4)
0.22Æ0.11 (8)
2.2Æ1 (30)
1050Æ250 (2)
>3000 (2)
1100Æ300 (2)
680Æ190 (2)
1260Æ400 (4)
660Æ100 (2)
790Æ10 (2)
–(CH2)2-Ph
–(CH2)-CH-(CH3)2
–(CH2)-CH-(CH3)2
–(CH2)3-CH3
m-CH3
m-CH3
m-CH3
m-CH3
m-CH3
m-CH3
m-CH3
H
m-CH3
o-CH3
m-CH3
m-CH3
m-CH3
m-CH3
IAc
–CH2-Ph
1120Æ300 (2)
1410Æ300 (2)
IA
>3000 (2)
710Æ320 (6)
330Æ120 (6)
IA
–CH2-(1-naphthyl)
–(CH2)2-N-(CH3)2
–(CH2)2-(o-pyridyl)
–CH2-(o-pyridyl)
–CH2-(o-pyridyl)
–CH2-(o-pyridyl)
–CH2-(p-pyridyl)
–(CH2)2-Ph
–CH3
–CH3
–(CH2)2CH3
–(CH2)2CH3
–(CH2)2CH3
–CH2CH3
–CH2Ph
–(CH2)2OCH2CH3
1000Æ320 (6)
250Æ10 (2)
1170Æ300 (2)
0110Æ10 (4)
10Æ5 (8)
17
18
–(CH2)2OCH2CH3
19
Raloxifene
Estradiol
H
—
—
–(CH2)2CH(CH2–Ph)–(CH2)2–
—
—
—
—
3.5Æ1 (30)
aSee Table 1 figure.
bThe values for Ki were obtained from least squares fit of the concentration–response curves according to the equation Àb=Àb0/1+[L]/Ki where
b0=the counts bound in the absence of test compound and b=counts bound in the presence of test compound at concentration [L]Æstandard
deviation (number of determinations).
cIA=inactive at 10À5 M.
.
Scheme 1. Reagents and conditions: (a) dihydropyran, H2SO4 SiO2 (cat), CH2Cl2, rt, 15 min, 85–95%; (b) ArCHO, NaOH, EtOH, rt, 16 h, 75–90%;
(c) 2-(benzotriazol-1-yl)acetonitrile, R3R4NH, EtOH, reflux, 24–48 h, 20–60%; (d) 80% AcOH, 60 ꢀC, 18 h, 50–65%.

B. R. Henke et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1939–1942
1941
adapted to
a
solution-phase, parallel-synthesis
tivity for binding to ERa versus ERb. Only a limited set
of substituents on the two phenyl rings were evaluated
in this initial compound set, with R1 and R2 being lim-
ited to –H, –CH3, and –OH. Placement of the hydroxyl
group in the para position of the 6-phenyl ring was
important for activity, as all analogues with a meta-
phenol were inactive at both ER subtypes (data not
shown). A methyl group meta to the phenol (R1=m-
CH3) increased the affinity modestly for both ER sub-
types versus the corresponding unsubstituted phenol
(compare entries 1 vs 4, 5 vs 6 in Table 1). This
increased affinity may be due to the methyl group’s
occupation of a hydrophobic pocket in both ERa and
ERb corresponding to the 6-position of the B-ring in
estradiol and, as evident in Figure 2, the sulfur in the
thiophene ring of raloxifene. Placement of this methyl
group ortho to the phenol led to a drastic loss of binding
affinity at both receptor subtypes (cf. 14 vs 15, Table 1).
Methyl substitution at the ortho position of the 4-phenyl
ring did not provide any substantial change in affinity
over the unsubstituted analogues (Table 1, 1 vs 3).
However, addition of a second phenol moiety either
in the para (2) or meta position (data not shown) of
the 4-phenyl ring led to a significant loss of receptor
binding affinity. This result might seem surprising
since addition of a second phenolic group tends to
increase ER binding affinity in many nonsteroidal
estrogens. However, our modeling (Fig. 2) suggests
that the 4-phenyl and 6-phenyl groups in our template
span a greater distance compared with the correspond-
ing aryl groups of raloxifene. Thus, placing substituents
at the meta and para positions of the 4-phenyl ring may
not be tolerated sterically within the binding pocket.
approach. Parallel chromatographic purification on
silica gel was done after chalcone formation and then
again after deprotection to afford the final target com-
pounds. Compound identity was established both by
mass spectrometry and 1H NMR. Compound purity
was assessed by HPLC analysis, with all reported
compounds displaying >90% purity (data not shown).
Compounds were tested for their ability to bind to ERa
and ERb via a scintillation proximity assay (SPA) using
a bacterial lysate containing overexpressed GST-hERa
or GST-hERb ligand binding domain. Yttrium silicate
SPA beads were suspended in assay buffer (10 mM
K2HPO4, 10 mM KH2PO4, 2mM EDTA, 50 mM NaCl,
1 mM DTT, 2mM CHAPS, 10% glycerol) and dis-
pensed at 0.5 mg/well. Lysates containing GST-hERa or
GST-hERb were diluted in assay buffer and added to
plates to give a final concentration of ꢁ0.15–0.2 mg
protein with a final assay volume of 100 mL. Test com-
pounds were dissolved in DMSO, serially diluted in
assay buffer and added to the wells in 10 mL aliquots.
1 nM [3H]17b-estradiol was then added and the plates
were shaken for 2h before radioactivity was counted.
The de novo design of this template originated by com-
parison of this scaffold to the known pharmacophore
and structural requirements for binding to the ER,
coupled with the assumption that a range of analogues
could be readily produced in parallel fashion using the
chemistry established by Katritzky.13 As part of the
template design process, we docked representative com-
pounds from this series into the published crystal struc-
ture of ERa bound with raloxifene8 using an in-house
molecular modeling package.14 Figure 2illustrates the
overlay of raloxifene and compound 19 obtained from
this docking procedure. Table 1 summarizes the binding
affinity of compounds 1–19 from this series to ERa and
ERb and serves as a brief summary of the structure–
activity relationships within the series. Compounds
within this set displayed a modest (2- to 5-fold) selec-
Investigation into the substituents on the 2-amino group
revealed that sterically large, nonpolar groups were
necessary to achieve good receptor affinity. The affinity
of these compounds may be due to their interactions
with residues along a large hydrophobic tunnel present
in the ligand-occupied antagonist conformations of
ERa and ERb.9 Either alkyl or aralkyl groups at R3 and
R4 provided compounds with submicromolar affinity.
Increasing chain length and bulk at either R3 (Table 1,
cf. entries 12 vs 14, 4 vs 17) or R4 (Table 1, cf. entries 6
vs 7, 4 vs 9) led to modest increases in receptor binding
affinity. Even the 1-naphthyl derivative 10 showed sub-
micromolar affinity at ERa. Small cyclic hydrocarbons
such as the pyrrolidine derivative 8 were inactive. Place-
ment of polar groups within R3 and R4 was detrimental to
binding affinity. For example, compound 11, which con-
tains a basic nitrogen in this region, lost all affinity for the
ER, and compound 18, containing two alkyl ethers, dis-
played only weak affinity for both ER subtypes. Replace-
ment of the phenyl group in R4 with pyridyl also led to a
loss in receptor affinity (Table 1, cf. entries 4 vs 12 and 9 vs
16). Interestingly, this potency loss can be recovered by
increasing the steric bulk of the R3 substituent, as com-
pounds 13 and 14 are among the most potent compounds
tested. Conformationally constraining the phenethyl moi-
ety in R4 into a 4-benzyl piperazine system afforded com-
pound 19, which is the most potent analogue within this
series at both receptor subtypes, with a Ki=20 nM at ERa
and a Ki=110 nM at ERb.
Figure 2. Overlay of raloxifene (in green) and compound 19 (in
orange) based on the docking of compound 19 into the crystal struc-
ture of raloxifene bound to ERa.

1942
B. R. Henke et al. / Bioorg. Med. Chem. Lett. 11 (2001) 1939–1942
In order to assess the functional profile of these novel
ER ligands we tested the transcriptional activity of
compound 4, as a representative of this series, in a
transient transfection assay utilizing a human breast
carcinoma cell line (T47D). T47D cells were transfected
with expression vectors containing full length hERa or
hERb, b-galactosidase, and an estrogen-responsive
reporter gene construct consisting of two copies of an
estrogen receptor response element, the estrogen-
responsive HSV tk promoter, and a SPAP reporter
gene. Alkaline phosphatase activity was corrected for
transfection efficiency using b-galactosidase activity as
an internal standard. Drug dilutions were prepared in
phenol red-free DMEM/F-12with 15 mM HEPES buf-
fer supplemented with 10% charcoal-stripped, delipi-
dated calf serum. Drug-treated cells were incubated for
24 h, after which the medium was sampled and assayed
for b-galactosidase and SPAP activity. Compound 4
profiled as a functional antagonist at ERb, fully sup-
pressing estradiol-stimulated transcriptional activity
with an IC50=160 nM (n=2). Compound 4 was a weak
partial agonist at ERa, with a maximal stimulation of
16% relative to estradiol and an EC50=30 nM (n=2).
The partial agonist nature of compound 4 at ERa was
confirmed by demonstrating that 4 was able to antag-
onize the effects of 17-b estradiol to 65% of its maximal
efficacy, with an IC50=30 nM (n=2). Compound 4 was
roughly 5-fold more potent in this assay than in the
binding assay, which is within reasonable agreement.
The modest selectivity for the ERa subtype observed in
this cell-based functional assay (b/a=5.3) is also in
good agreement with the subtype selectivity observed in
the binding assay (b/a=4.3).
These compounds show a modest preference for binding
to the ERa subtype, with the most potent compound in
this series (19) having a Ki=20 nM at ERa. Compound
4 from this series profiled as a full antagonist at ERb and
a weak partial agonist at ERa in a cell-based functional
assay. These ligands may have utility in the treatment of
various diseases associated with estrogen loss.
References and Notes
1. Cosman, F.; Lindsay, R. Endocr. Rev. 1999, 20, 418.
2. Lin, X.; Huebner, V. Curr. Opin. Drug Discovery Dev. 2000,
3, 383.
3. Grese, T. A.; Dodge, J. A. Curr. Pharm. Des. 1998, 4, 71.
4. Kuiper, G. G. J. M.; Enmark, E.; Pelto-Huikko, M.; Nils-
son, S.; Gustafsson, J.-A. Proc. Natl. Acad. Sci. U.S.A. 1996,
93, 5925.
5. Kuiper, G. G. J. M.; Carlsson, B.; Grandien, K.; Enmark,
E.; Haeggblad, J.; Nilsson, S.; Gustafsson, J.-A. Endocrinology
1997, 138, 863.
6. Couse, J. F.; Korach, K. S. Endocr. Rev. 1999, 20, 358.
7. Anstead, G. M.; Carlson, K. E.; Katzenellenbogen, J. A.
Steroids 1997, 62, 268.
8. Brzozowski, A. M.; Pike, A. C. W.; Dauter, Z.; Hubbard,
R. E.; Bonn, T.; Engstrom, O.; Ohman, L.; Greene, G. L.;
Gustafsson, J.-A.; Carlquist, M. Nature 1997, 389, 753.
9. Wurtz, J.-M.; Egner, U.; Heinrich, N.; Moras, D.; Mueller-
Fahrnow, A. J. Med. Chem. 1998, 41, 1803.
10. Pike, A. C. W.; Brzozowski, A. M.; Hubbard, R. E.;
Bonn, T.; Thorsell, A.-G.; Engstrom, O.; Ljunggren, J.; Gus-
tafsson, J.-A.; Carlquist, M. EMBO J. 1999, 18, 4608.
11. Chavez, F.; Godinez, R. Synth. Commun. 1992, 22, 159.
12. Wattanasin, S.; Murphy, W. S. Synthesis 1980, 647.
13. Katritzky, A. R.; Belyakov, S. A.; Sorochinsky, A. E.;
Henderson, S. A.; Chen, J. J. Org. Chem. 1997, 62, 6210.
14. Lambert, M. H. In Practical Application of Computer-
Aided Drug Design; Charifson, P. S., Ed.; Marcel Dekker:
New York, 1997; Chapter 8, pp 243–303.
In summary, utilizing solution-phase, parallel-synthesis
techniques we have synthesized a series of 2-amino-4,6-
diarylpyridines that are novel ligands for ERa and ERb.