Exploration of the bicyclo[3.3.1]nonane system as a template for the development of new ligands for the estrogen receptor

Article abstract of DOI:10.1016/j.bmcl.2003.08.061

Three novel structural motifs based on a bicyclo [3.3.1]nonane template were examined as new ligands for estrogen receptor (ER). Type III compounds emerged as the most promising leads for developing high-affinity ER ligands, but they showed little selecti

Full text of DOI:10.1016/j.bmcl.2003.08.061

Bioorganic & Medicinal Chemistry Letters 13 (2003) 4485–4488
Exploration of the Bicyclo[3.3.1]nonane System as a Template
for the Development of New Ligands for the Estrogen Receptor
Rajeev S. Muthyala, Kathryn E. Carlson and John A. Katzenellenbogen*
Department of Chemistry, University of Illinois, 600 S Mathews Avenue, Urbana, IL 61801, USA
Received 30 July 2003; revised 14 August 2003; accepted 27 August 2003
Abstract—Three novel structural motifs based on a bicyclo [3.3.1]nonane template were examined as new ligands for estrogen
receptor (ER). Type III compounds emerged as the most promising leads for developing high-affinity ER ligands, but they showed
little selectivity for either ER subtype. Type II compounds, on the other hand, despite their lower affinity, exhibited significant ERb
binding selectivity.
# 2003 Elsevier Ltd. All rights reserved.
The estrogen receptor (ER) is a ligand-dependent tran-
scription factor whose native hormonal ligand is estra-
diol (E2). The actions of estrogens are mediated through
the ER of which there are two subtypes, ERa and
ERb.1,2 Based on crystallographic and ligand analogue
studies,3À7 we now have a reasonable understanding of
the structural requirements for a ligand to bind to ER:
The ER pharmacophore basically consists of (1) an A-
ring phenol, (2) a second hydroxyl group or phenol
placed approximately 11 A from the A-ring phenol to
mimic the 17b-hydroxy group of E2, and (3) a central
structurally variable hydrophobic core which mimics
the B and C rings of estradiol.8
the ligand, particularly near the middle of the ligand
(i.e., near the B and C rings of E2).3 By incorporating a
hydrophobic bicyclic core, we hoped to exploit this
unfilled space in the ER binding pocket, and thereby,
potentially, to gain binding affinity or ER subtype
selectivity.
A brief survey of the various available bicyclic cores
suggested that a bicyclo[3.3.1]nonane system would be
an appropriate three-dimensional hydrophobic core
element, because its overall dimensions and carbon
atom count closely match those of the B and C rings of
E2. In addition, we were encouraged by the fact that
several synthetic routes are available to prepare bicy-
clo[3.3.1]nonanes.9 Therefore, we proposed the three
different structural motifs shown below, each containing
a bicyclo[3.3.1]nonane core, as possible E2 mimics, and
we have conducted a limited structure–activity relation-
ship (SAR) study of each type.
As part of our long-term interest in ER ligands, we
undertook an exploratory study aimed at preparing new
ligands wherein the central hydrophobic core had over-
all a more three-dimensional topology. This design
strategy was based on an examination of the binding
pockets of both ERa and ERb that shows there is sub-
stantial unoccupied space above and below the plane of
Type I ligands, inspired by Troger’s base,10 consist of a
central bicyclo[3.3.1]nonane core fused to phenol
groups. In this design, we reasoned that the periphery of
the bridged bicyclic core would mimic the dimensions of
the B and C rings of estradiol, whereas the bridging
methylene group would provide steric bulk above (or
*Corresponding author. Tel.: +1-217-333-6310, fax: +1-217-333-
7325; e-mail: jkatzene@uiuc.edu
0960-894X/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2003.08.061

4486
R. S. Muthyala et al. / Bioorg. Med. Chem. Lett. 13 (2003) 4485–4488
below) the rest of the ligand. The structural motifs in
the remaining two types of ligands are such that the
phenols are attached to the bicyclic core at various
positions, either directly (Type II), or by an sp2-hybri-
dized carbon (Type III). The latter motif, with two
phenols attached to the bridgehead as a diarylmethyli-
dine unit, was inspired by the high affinity ER ligand,
cyclofenil.
The Troger’s base analogue, compound 1, with bridge-
head nitrogen atoms and a bicyclo[3.3.1]nonane core,
was synthesized from 4-methoxyaniline as described10
and was deprotected with BBr3. Type I ligands with
bridgehead carbon atoms were synthesized by the gen-
eral route shown in Scheme 1 using methodology
described earlier.11
Scheme 2.
subsequent deoxygenation step, hydride donation from
triethylsilane again occurs from the exo face, thus lead-
ing to an endo stereochemistry for the two methyl
groups on the periphery of the bicyclic core. It is
important to note that all chiral ligands described in this
study were obtained and used as racemic mixtures.
In the first step, a substituted phenylacetonitrile was
alkylated with diiodomethane to give a mixture of meso
and dl-diphenyl glutaronitriles (4 and 5). This mixture
was hydrolyzed to furnish a diasteromeric mixture of
diphenylglutaric acids (6 and 7), which were subsequently
subjected to a double intramolecular Friedel–Crafts
acylation using polyphosphoric acid or methanesulfonic
acid/P2O5,12 furnishing the cyclized products 8 and 9.
These latter compounds could be further transformed
into additional analogues, as shown in Scheme 2, using
the keto groups as reactive handles.
The synthesis of Type II ligands was accomplished
starting from commercially available bicyclic ketones 18
and 22, using the synthetic transformations shown in
Scheme 3. A methyl group at C-9 (compound 21) was
introduced by nucleophilic substitution on tertiary
benzylic alcohol 2 with trimethylaluminum, used in
conjunction with boron trifluoride etherate.14 The two
phenyl groups in 24 were also determined to be endo,
which can be rationalized in the same manner as for
compounds 16 and 17 above. The introduction of the
second phenol group at C-9 (compound 25) was
accomplished using established methods.15
For example, methyl substituents on the periphery of
the bicyclic core could be introduced by addition of
methyl magnesium bromide to the bis-keto compounds
8 and 9. This was followed by deoxygenation using a
triethylsilane–trifluoroacetic acid mixture.13 The endo
stereochemistry of the methyl groups in 16 and 17 was
confirmed by X-ray crystallography. The exclusive for-
mation of the endo isomers can be rationalized by
assuming that the initial addition of the Grignard
reagent occurs from the less hindered exo face. In the
Scheme 1.
Scheme 3.

R. S. Muthyala et al. / Bioorg. Med. Chem. Lett. 13 (2003) 4485–4488
4487
Type III ligands were synthesized by crossed McMurry
coupling, as shown in Scheme 4. The latter reaction was
particularly efficient when a diaryl ketone was one of
the reactants.16
affinity (RBA) values, where estradiol has an affinity of
100%.
The binding affinity of the Troger’s base analogue
(compound 1) to the estrogen receptor was below the
limits of our assay. Initially, it was not clear whether
the low binding affinity of 1 was due to its highly con-
cave nature or to the presence of the bridgehead nitro-
gen atoms. To examine this, we synthesized the
carbocyclic analogues in which the bridgehead nitrogen
atoms were replaced by carbon atoms (compounds 12
and 13).
Following the synthesis of different ligands, we exam-
ined their binding to both ERa and ERb. The binding
affinities of all the ligands prepared in this study are
given in Table 1 and are expressed as relative binding
The significantly increased binding affinity of these lat-
ter compounds suggests that the polar bridgehead
nitrogen atoms in 1 are responsible for its immeasurably
low binding affinity.
Scheme 4.
Table 1. Relative binding affinity (RBA) of bicyclic non-steroidal analogues for estrogen receptors a and b. The RBA values were determined in a
competitive radiometric binding assay described earliera,17,18
No.
Compound
ERa
ERb
b/a
No.
Compound
ERa
ERb
b/a
1
<0.005
<0.005
—
20
0.13Æ0.03
1.33Æ0.26
10
12
13
0.033Æ0.005
0.22Æ0.011
0.072Æ0.004
0.77Æ0.15
2
21
30
0.32Æ08.90
0.39Æ0.006
Æ028
3.5
1.81Æ0.15
4.6
6.6
16
17
28
29
0.007Æ0.0
0.025Æ0.003
3.6
2.7
3.5
4.4
24
25
26
27
0.44Æ0.02
0.31Æ0.02
1.71Æ0.45
513Æ17
2.9Æ0.2
0.25Æ0.02
4.58Æ1.3
301Æ54
0.004
0.011
0.8
0.006Æ0.001
0.005Æ0.001
0.021Æ0.006
0.022Æ0.004
2.7
0.59
aThe RBA of estradiol is 100, values represent the averageÆrange or SD of 2–3 independent determinations.

4488
R. S. Muthyala et al. / Bioorg. Med. Chem. Lett. 13 (2003) 4485–4488
For Type I ligands, it is intriguing to note that the
binding affinity depends on the position of the bridge.
For compound 13, in which the bridge is para to the
hydroxyl group, the binding affinity is higher than in
compound 12, in which the bridge is meta to the phe-
nolic hydroxyl group. Another feature that is readily
apparent from the SAR study for Type I ligands is that
ER does not seem to tolerate any substituent, electron
donating or withdrawing, on the periphery of the
bicyclic core. For example, note the significantly
reduced binding affinities of both 16 and 28, compared
to 12 and 13, respectively.
selectivity. It remains to be seen whether this selectivity
and affinity can be further enhanced by the addition of a
polar functionality on the bicyclic core to mimic the
distal ring-OH group of E2. Studies in this direction will
be the subject of future investigations and will be
reported in due course.
Acknowledgements
This research was supported by grants from the
National Institutes of Health (PHS 5R37 DK15556).
Compounds containing phenols directly attached to the
bicyclic core (Type II ligands) exhibit higher affinity
toward ER compared to Type I ligands. Presumably,
this is in part due to their increased flexibility. What is
noteworthy, however, is that some of the Type II
ligands (20 and 21) have significant ERb affinity selec-
tivity. Monophenol 20 exhibits a 15-fold selectivity for
ERb which is further increased by the introduction of a
methyl group at C-9 as in 21. The ERb binding selec-
tivity of 21 is comparable to some of the recently
reported triazines.19 Bisphenols 24 and 25 have lower
binding affinities and exhibit little or no binding selec-
tivity for ERb. Preliminary data shows that compound
21 does not exhibit substantial selectivity for ERb in cell
based transfection assays, either in terms of efficacy or
potency (S. Sheng and B. S. Katzenellenbogen, unpub-
lished). During the preparation of this manuscript, we
became aware of another report pertaining to the
synthesis of non-steroidal ligands having bicyclo[3.3.1]-
nonane cores.20 However, the binding affinities and the
ERb, selectivities of the ligands described in that report
are lower than those of compound 21.
References andNotes
1. Katzenellenbogen, J. A.; O’Malley, B. W.; Katzenellenbo-
gen, B. S. Mol. Endocrinol. 1996, 10, 119.
2. Katzenellenbogen, B. S.; Katzenellenbogen, J. A. Breast
Cancer Res. 2000, 2, 335.
3. Brzozowski, A. M.; Pike, A. C.; Dauter, Z.; Hubbard,
R. E.; Bonn, T.; Engstrom, O.; Ohman, L.; Greene, G. L.;
Gustafsson, J. A.; Carlquist, M. Nature 1997, 389, 753.
4. Shiau, A. K.; Barstad, D.; Loria, P. M.; Cheng, L.; Kush-
ner, P. J.; Agard, D. A.; Greene, G. L. Cell 1998, 95, 927.
5. Anstead, G. M.; Carlson, K. E.; Katzenellenbogen, J. A.
Steroids 1997, 62, 268.
6. Gao, H.; Katzenellenbogen, J. A.; Garg, R.; Hansch, C.
Chem. Rev. 1999, 99, 723.
7. Lin, X.; Huebner, V. Curr. Opin. Drug Discov. Dev. 2000, 3,
383.
8. Fink, B. E.; Mortensen, D. S.; Stauffer, S. R.; Aron, Z. D.;
Katzenellenbogen, J. A. Chem. Biol. 1999, 6, 205.
9. Butkus, E. Synlett 2001, 1827.
10. Spielman, M. A. J. Am. Chem. Soc. 1935, 57, 583.
11. Ogura, F.; Nakao, A.; Nakagawa, M. Bull. Chem. Soc.
Jpn. 1979, 52, 1165.
12. Eaton, P. E.; Carlson, G. R.; Lee, J. T. J. Org. Chem.
1973, 38, 4071.
13. Doyle, M. P.; McOsker, C. C.; Ball, N.; West, C. T. J.
Org. Chem. 1977, 42, 1922.
14. Uemura, M.; Isobe, K.; Hayashi, Y. Tetrahedron Lett.
1985, 26, 767.
15. Liaw, D.-J.; Liaw, B.-Y.; Chung, C.-Y. Acta Polym. 1999,
50, 135.
16. McMurry, J. E.; Krepski, L. R. J. Org. Chem. 1976, 41,
3929.
17. Katzenellenbogen, J. A.; Johnson, H. J., Jr.; Myers, H. N.
Biochemistry 1973, 12, 4085.
18. Carlson, K. E.; Choi, I.; Gee, A.; Katzenellenbogen, B. S.;
Katzenellenbogen, J. A. Biochemistry 1997, 36, 14897.
19. Henke, B. R.; Consler, T. G.; Go, N.; Hale, R. L.; Hoh-
man, D. R.; Jones, S. A.; Lu, A. T.; Moore, L. B.; Moore,
J. T.; Orband-Miller, L. A.; Robinett, R. G.; Shearin, J.;
Spearing, P. K.; Stewart, E. L.; Turnbull, P. S.; Weaver, S. L.;
Williams, S. P.; Wisely, G. B.; Lambert, M. H. J. Med. Chem.
2002, 45, 5492.
Of the various ligands investigated in this study, type III
ligands exhibit the highest affinity for ER. Remarkably,
unlike a similar substitution in the type II series, the
replacement of a methyl group in 26 with another phe-
nol leads to a very high affinity ligand (compound 27),
having an RBA of 500% of that of E2. In fact, this
binding affinity is almost twice that of the well-known
estrogen cyclofenil (RBA: ERa=68 and ERb=334),
which is similar in all respects to 27 except for the three
carbon bridge. This indicates that for certain types of
structures, the binding affinity of a ligand can be
increased simply by an increase in the number of
hydrophobic interactions through the introduction of a
bicyclic core. Interestingly, Type III ligands derived
from other three-dimensional cores having a larger or
smaller number of carbon atoms showed reduced affi-
nity for ER. We have conducted and reported elsewhere
a detailed structure–activity study of Type III ligands.21
20. Sibley, R.; Hatoum-Mokdad, H.; Schoenleber, R.; Musza,
L.; Stirtan, W.; Marrero, D.; Carley, W.; Xiao, H.; Dumas, J.
Bioorg. Med. Chem. Lett. 2003, 13, 1919.
21. Muthyala, R. S.; Sheng, S.; Carlson, K. E.; Katze-
nellenbogen, B. S.; Katzenellenbogen, J. A. J. Med. Chem.
2003, 46, 1589.
In conclusion, by examining three novel structural
motifs as E2 mimics, we identified Type III ligands as
lead compounds for developing high-affinity ER
ligands. Type II ligands show significant ERb affinity