Comparison of 2-phenylspiroindenes and 2-phenylspiroindenediones as estrogen receptor ligands - Modeling and binding data don't agree!

Article abstract of DOI:10.1016/j.bmcl.2003.12.058

A series of 2-phenylspiroindenediones was prepared. The spiroindenediones were found to be less active than the corresponding spiroindenes as estrogen receptor ligands and failed to demonstrate the receptor subtype selectivity that had been anticipated from molecular modeling.

Full text of DOI:10.1016/j.bmcl.2003.12.058

Bioorganic & Medicinal Chemistry Letters 14 (2004) 1317–1321
Comparison of 2-phenylspiroindenes and 2-phenylspiroindenediones
as estrogen receptor ligands—modeling and binding data
don’t agree!
Timothy A. Blizzard,* Ralph T. Mosley, Elizabeth T. Birzin, Wanda Chan
and Milton L. Hammond
Merck Research Laboratories, RY800-B116, PO Box 2000, Rahway NJ 07065USA
Received 7August 2003; accepted 5 December 2003
Abstract—A series of 2-phenylspiroindenediones was prepared. The spiroindenediones were found to be less active than the
corresponding spiroindenes as estrogen receptor ligands and failed to demonstrate the receptor subtype selectivity that had been
anticipated from molecular modeling.
# 2004 Elsevier Ltd. All rights reserved.
Hormone replacement therapy (HRT) is widely used to
treat a variety of conditions, such as hot flashes and
osteoporosis, resulting from estrogen deficiency in post-
menopausal women.1 Although HRT is very effective, it
is also associated with some serious side-effects, includ-
ing blood clots and increased risk of cancer. The selec-
tive estrogen receptor modulators (SERMs) have
generated considerable interest due to their potential
ability to provide the benefits of estrogen without the
associated liabilities.2 The clinical utility of SERMs is
exemplified by the use of tamoxifen3 for breast cancer
and raloxifene4 for osteoporosis.4d,e The recent dis-
covery of a second ER receptor subtype (ERb)5 suggests
the possibility of developing subtype selective SERMs.
Most of the current SERMs are ‘balanced’ compounds,
that is, they bind both receptor subtypes equally,
although several novel classes of subtype selective com-
pounds have recently been described.6 Due to the vari-
able tissue distribution of the two receptor subtypes, it is
possible that selectivity for one receptor over the other
might confer clinical advantages.
in ERa (Fig. 1) suggested that addition of a carbonyl
oxygen at C-30 of the spiroindene system of 1 would
result in substantial selectivity for the ERa receptor due
to negative steric and electronic clashes with the Met
293 side-chain of ERb (analogous to Leu 384 of ERa).
Although the carbonyl at C-10 of 2 was predicted to have
no effect on selectivity and would merely mimic the car-
bonyl found in raloxifene, its presence simplified the
synthesis somewhat. We thus targeted spiroindenedione
2 for synthesis with the expectation that 2 would be
both potent and selective for ERa.
Recently, we reported a series 2-phenylspiroindenes
(e.g., 1) as estrogen receptor ligands.7 The spiroindenes
were very potent but were not selective for either ERa
or ERb. Molecular modelling studies of 1 and 2 docked
We initially attempted to prepare 2 by modifying our
synthetic route to 1.7 Synthesis of the acid cyclization
precursor
5
proceeded uneventfully (Scheme 1).8
However, cyclization of 5 to the advanced ketone inter-
mediate 6 was plagued by an unavoidable fragmenta-
tion reaction that regenerated dione 3 along with many
* Corresponding author. Fax: +1-732-594-9556; e-mail: tim_blizzard@
merck.com
0960-894X/$ - see front matter # 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2003.12.058

1318
T. A. Blizzard et al. / Bioorg. Med. Chem. Lett. 14 (2004) 1317–1321
enolate onto an olefin (Scheme 2). The key cyclization
step was based on the successful application of a similar
reaction by Rao et al. in their studies toward freder-
icamycin A.9 The requisite precursors, triflate 710 and
stannane 8,11 were easily prepared from commercially
available starting materials. Palladium mediated cou-
pling of 7 and 8 using conditions developed by Nemoto
et al.10 for vinylation of 8 afforded aldehyde 9 in very
good yield. Condensation of 9 with phthalide 1012 under
the conditions used by Shapiro et al.13 for a simpler sys-
tem afforded the key cyclization intermediate 11 in 55%
yield.14 Palladium mediated cyclization of 11 using the
Rao conditions9 cleanly afforded the desired spiro-
indenedione 12 in 50% yield.15 Our initial synthetic
plan had been to selectively deblock the methyl ether
para to the ketone of 12 (to afford 13) then attach the
piperidinylethoxy side chain and deprotect the other
phenols to afford 2. In practice, it was not possible to
selectively cleave the desired methyl ether, probably due
to competing nucleophilic attack at one of the ketone
carbonyls. It was, however, possible to remove the other
methyl ethers with boron tribromide to afford 14 for
biological evaluation.
Figure 1. Spiroindenes 1 and 2 were modeled within the context of
human ERa using the co-crystallized antagonist raloxifene as a start-
ing point.
other products. Under some conditions, it was possible
to isolate small amounts (ꢀ10%) of 6 but the reaction
was extremely messy and this was an unacceptably low
yield at this relatively early stage of the synthesis so this
approach was abandoned.
Several alternative approaches involving bis acylation of
an indene anion or oxidation of a spiroindene were
similarly unsuccessful, possibly due to product instabil-
ity. Finally, we turned to an approach based on a pal-
ladium mediated intramolecular cyclization of a dione
Unfortunately, 14 was substantially less active in an
estrogen receptor binding assay (IC50s: human ERa=137
nM; ERb=153 nM) than we had hoped and, more
importantly, it was not selective for ERa. Nevertheless,
Scheme 1. Conditions: (i) K2CO3, DMF, 50 ꢁC, 15 m, 94%; (ii) KOH, MeOH, H2O, 18 h, 58%; (iii) PCl5, CH2Cl2, 0 ꢁC, 1 h; (iv) AlCl3, CH2Cl2,
0 ꢁC, 45 m, ꢀ10% from 5.
Scheme 2. Conditions: (i) Pd(PPh3)4, LiCl, THF, reflux, 6 h, 84%; (ii) 4.5 equiv NaOMe, MeOH, ethyl propionate, 82 ꢁC, 50 m, 55%; (iii) NaH,
Pd(OAc)2, DMF, rt, 2 h, 50%; (iv) NaSMe, DMF, decomposed; (v) BBr3, CH2Cl2, rt, 75 m, 32%.

T. A. Blizzard et al. / Bioorg. Med. Chem. Lett. 14 (2004) 1317–1321
1319
since we were finally so close to accomplishing the goal
of synthesizing 2, we decided to complete the synthesis
and hope that the addition of the side chain would
improve the potency and selectivity.
Interestingly, 20 was highly susceptible to air oxidation
and cyclization to afford primarily the cyclic hydroper-
oxide 23 along with small amounts of ketone 24 and
other unidentified minor products. Indeed, when the
condensation of 19 and 9 was run in air rather than
under Argon, hydroperoxide 23 was isolated as the
main product along with ketone 24 and several uni-
dentified minor byproducts. The structure of 23 was
confirmed by extensive NMR studies and by its conver-
sion to a mixture of 22 and 24 upon treatment with p-
toluenesulfonic acid in hot benzene.
An initial attempt to complete the synthesis of 2 by
substituting a MEM protecting group for the trouble-
some methyl ether failed when the MEM group also
proved to be impossible to remove without destroying
the spiroindenedione system. In the successful approach
the piperidinylethoxy side chain was incorporated in the
beginning of the synthesis, thus eliminating the need for
a protecting group altogether (Scheme 3). In this
sequence, the cyclization precursor 17 was not isolated
and purified but was instead cyclized immediately to
form the penultimate intermediate 18 which was depro-
tected to afford the long sought spiroindenedione 2.
Unfortunately, 2 was also only moderately active in the
binding assay and was not selective for ERa (Fig. 2).16
With the synthetic methodology for synthesizing spiro-
indenediones in hand, two additional analogues were
quickly prepared for direct comparison with previously
prepared spiroindenes. Thus, reaction of phthalide 19
with aldehyde 9 afforded the cyclization intermediate 20
in 80% crude yield (Scheme 4). Palladium mediated
cyclization of crude 20 afforded the penultimate inter-
mediate 21 which was deblocked uneventfully to afford
the spiroindenedione 22.
The final spiroindenedione to be prepared was the meta
isomer 29. The requisite intermediate aldehyde 26 was
prepared from stannane 2517 using chemistry analogous
to that used to prepare 9. Application of our stan-
dard protocol to aldehyde 26 afforded the cyclization
precursor 27 in 95% crude yield. Cyclization and
deprotection proceeded uneventfully to afford 29
(Scheme 5).
Scheme 3. Conditions: (i) N-chloroethylpiperidine, K2CO3, MeCN, 40 ꢁC, 37h, 66%; (ii) 9, 2.2 equiv NaOMe, MeOH, ethyl propionate, 82 ꢁC, 24 h;
(iii) Pd(OAc)2, DMF, rt, 1 h, ꢀ10% from 16; (iv) HCl, ether, then BBr3, CH2Cl2, rt, 75 m, 27%.
Scheme 4. Conditions: (i) 9, 4.5 equiv NaOMe, MeOH, ethyl propionate, Argon, 82 ꢁC, 24 h 80% crude; (ii) NaH, Pd(OAc)2, DMF, rt, 90 m, 30%;
(iii) BBr3, CH2Cl2, rt, 75 m, 58%.

1320
T. A. Blizzard et al. / Bioorg. Med. Chem. Lett. 14 (2004) 1317–1321
Scheme 5. Conditions: (i) 7, Pd(PPh3)4, LiCl, THF, reflux, 7h, 46%; (ii) 4.5 equiv NaOMe, MeOH, ethyl propionate, Argon, 82 ꢁC, 90 m, 95%
crude; (iii) NaH, Pd(OAc)2, DMF, rt, 2 h, 25%; (iv) BBr3, CH2Cl2, rt, 75 m, 78%.
An alternative cyclization method was also briefly
explored. Thus, treatment of intermediate 11 with N-
iodosuccinimide (NIS) in the presence of base afforded
a low yield of the desired spiroindenedione 12 along
with a substantial amount of the cyclized iodide 30.
Presumably, 30 could be readily converted to 12
although this was not actually done. This approach
therefore appears to provide a viable alternative to the
palladium-mediated cyclization for preparing spiro-
indenediones. However, this promising initial result
was not followed up since the disappointing results from
the evaluation of our spiroindenediones in the ER
binding assay did not merit further work in this area
(Scheme 6).
groups apparently generate an unfavorable interaction
with both estrogen receptors rather than with just the
beta receptor. Interestingly, this effect is most pro-
nounced in the fully elaborated analogue 2. This sug-
gests that the range of motion available to the side chain
may be more important than was previously realized.
Also noteworthy is the observation that 2, which incor-
porates the raloxifene side chain, is slightly less active
than 22, which does not have the side chain. This con-
trasts with the corresponding spiroindene pair (1 and
31) wherein introduction of the side chain results in a
slight increase in activity. As with the spiroindenes, the
analogue with the para hydroxy group in the pendant
phenol (22) is more active than the compound with a
meta hydroxy in this ring (29).
The binding data for the spiroindenediones and the
corresponding spiroindenes are summarized in Figure
2.16 Only those compounds which have been prepared in
both series are included in Figure 2; analogue 14 (see
binding data above) was omitted since the correspond-
ing spiroindene analogue was not prepared. In all cases,
introduction of the carbonyl groups at C-10 and C-30 of
the spiroindene system resulted in a substantial decrease
in potency with no gain in selectivity. In contrast to the
molecular modeling prediction, the additional carbonyl
It is clear that the spiroindenediones are consistently
less potent estrogen receptor binders than the spiro-
indenes and are not selective for either estrogen recep-
tor. In this series, molecular modeling was clearly not
successful in predicting the biological consequences of a
relatively small chemical structure change despite the
availability of a crystal structure of the receptor with a
structurally related ligand in the binding site. It remains
to be seen whether the introduction of other substituents
Scheme 6. Conditions: (i) NaH, NIS, THF, rt, 5 h.
Figure 2. Comparison of estrogen receptor binding affinities (IC50).16

T. A. Blizzard et al. / Bioorg. Med. Chem. Lett. 14 (2004) 1317–1321
1321
at these positions can impart the desired selectivity
while retaining potency.
P.; Bischoff, S. F.; Nozulak, J.; Feuerbach, D.; Floer-
sheim, P. J. Med. Chem. 2002, 45, 1399. (d) Myers, M. J.;
Sun, J.; Carlson, K. E.; Marriner, G. A.; Katzenellenbo-
gen, B. S.; Katzenellenbogen, J. A. J. Med. Chem. 2001,
44, 4230. (e) Mortensen, D. J.; Rodriguez, A. L.; Carlson,
K. E.; Sun, J.; Katzenellenbogen, B. S.; Katzenellenbo-
gen, J. A. J. Med. Chem. 2001, 44, 3838. (f) Myers, M. J.;
Sun, J.; Carlson, K. E.; Katzenellenbogen, B. S.; Katze-
nellenbogen, J. A. J. Med. Chem. 1999, 42, 2456.
References and notes
1. (a) Manson, J. E.; Martin, K. A. N. Engl. J. Med. 2001,
345, 34. (b) Keating, N. L.; Cleary, P. D.; Rossi, A. S.;
Zaslavsky, A. M.; Ayanian, J. Z. Ann. Intern. Med. 1999,
130, 545. (c) Kooistra, T.; Emeis, J. J. Curr. Opin. Chem.
Biol. 1999, 3, 495.
2. (a) Jordan, C. Scientific American 1998, 279, 60. (b)
Miller, C. P.; Komm, B. S. In Annual Reports In Medic-
inal Chemistry; Doherty, A. M., Ed.; Academic Press: San
Diego, 2001; Vol. 36. Ch. 15., p 149. (c) Park, W. C.;
Jordan, V. C. Trends Mol. Med. 2002, 8, 82. (d) Lonard,
D. M.; Smith, C. L. Steroids 2002, 67, 15. (e) Jordan, V. C.
J. Med. Chem. 2003, 46, 883. (f) Jordan, V. C. J. Med.
Chem. 2003, 46, 1081.
3. Jordan, V. C. Br. J. Pharmacol. 1993, 110, 507.
4. (a) Jones, C. D.; Jevnikar, M. G.; Pike, A. J.; Peters,
M. K.; Black, L. J.; Thompson, A. R.; Falcone, J. F.;
Clemens, J. A. J. Med. Chem. 1984, 27, 1057. (b) Griese,
T. A.; Pennington, L. D.; Sluka, J. P.; Adrian, M. D.;
Cole, H. W.; Fuson, T. R.; Magee, D. E.; Phillips, L.;
Rowley, E. R.; Shetler, P. K.; Short, L. L.; Venugopalan,
M.; Yang, N. N.; Sato, M.; Glasebrook, A. L.; Bryant,
H. U. J. Med. Chem. 1998, 41, 1272. (c) Griese, T. A.;
Sluka, J. P.; Bryant, H. U.; Cullinan, G. J.; Glasebrook,
A. L.; Jones, C. D.; Matsumoto, K.; Palkowitz, A. D.;
Sato, M.; Termine, J. D.; Winter, M. A.; Yang, N. N.;
Dodge, J. A. Proc. Natl. Acad. Sci. U.S.A. 1997, 94,
14105. (d) Sato, M.; Griese, T. A.; Dodge, J. A.; Bryant,
H. U.; Turner, C. H. J. Med. Chem. 1999, 42, 1. (e) Delmas,
P. D.; Bjarnson, N. H.; Mitlak, B. H.; Ravoux, A. C.;
Shah, A. S.; Huster, W. J.; Draper, M.; Christiansen, C.
N. Engl. J. Med. 1997, 337, 1641.
7. Blizzard, T. A.; Morgan, J. D.; Mosley, R. T.; Birzin,
E. T.; Frisch, K.; Rohrer, S. P.; Hammond, M. L. Bioorg.
Med. Chem. Lett. 2003, 13, 479.
8. All new compounds are racemic and were characterized
1
by LC-MS and 400, 500, or 600 MHz H NMR.
9. (a) Rao, A. V. R.; Reddy, D. R.; Rao, B. V. Ind. J. Chem.
1988, 27B, 1065. (b) Rao, A. V. R.; Rao, B. V.; Reddy,
D. R. Ind. J. Chem. 1991, 30B, 723.
10. Nemoto, H.; Miyata, J.; Yoshida, M.; Raku, N.; Fuku-
moto, K. J. Org. Chem. 1997, 62, 7850. Triflate 7 could be
stored in the freezer for more than a month without sig-
nificant decomposition..
11. Corey, E. J.; Wollenberg, R. H. J. Org. Chem. 1975, 40,
2265.
12. Prepared in 90% yield by alkylation of 6-hydroxy-
phthalide (prepared in turn from commercial 6-amino-
phthalide using the procedure reported by Vaughan et al.:
Vaughan, W. R.; Baird, S. L., Jr. J. Am. Chem. Soc. 1946,
68, 1314) with MeI and K2CO3 in MeCN at 35 ꢁC for 16 h.
13. Shapiro et al. J. Org. Chem. 1960, 25, 1860. The use of
ethyl propionate as the cosolvent proved to be critical for
the success of this reaction.
14. Note that in principle either 6-methoxyphthalide (10) or
5-methoxyphthalide (not shown) could be used in the
reaction with 9 to generate 11.
15. The structure of 12 was confirmed by extensive NMR
studies including full assignment of the 1H and 13C NMR
spectra using HMQC and HMBC experiments.
5. (a) Kuiper, G. G. J. M.; Enmark, E.; Pelto-Kuikko, M.;
Nilsson, S.; Gustafsson, J. A. Proc. Natl. Acad. Sci. 1996,
93, 5925. (b) Mosselman, S.; Polman, J.; Dijkema, R.
FEBS Lett. 1996, 392, 49.
6. (a) Ghosh, U.; Ganessunker, D.; Sattigeri, V. J.; Carlson,
K. E.; Mortensen, D. J.; Katzenellenbogen, B. S.; Katze-
nellenbogen, J. A. Bioorg. Med. Chem. 2003, 11, 629. (b)
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. (c) Schopfer, U.; Schoefter,
16. The IC50 values were generated in an estrogen receptor
ligand binding assay. This scintillation proximity asssay
was conducted in NEN Basic Flashplates using tritiated
estradiol and full length recombinant human ERa and
ERb proteins. Most of the data reflects a 3 h incubation
time; data marked with an * reflects a 4 h incubation.
Compounds were evaluated in duplicate in a single assay.
In our experience, this assay provides IC50 values that are
reproducible to within a factor of 2–3.
17. Prepared in 55% yield by reaction of 3-methoxy-
phenylacetylene with tributyltin hydride using a proce-
dure analogous to one described by Corey et al.: Corey,
E. J.; Wollenberg, R. H. J. Org. Chem. 1975, 40, 2265.