2402
C. P. Miller et al. / Bioorg. Med. Chem. Lett. 13 (2003) 2399–2403
Table 2. Gene regulation potency differences on ERa and ERb
endpoints
the non-selective binder 17b-estradiol. This magnification
of the binding selectivity observed for compound 6 relative
to 17b-estradiol could be explained by differential cellular
up take, metabolism, and so on across the two cell lines.
Alternatively, the differential transactivational potency
and efficacy maybe due to differential coregulatory
recruitment and utilization between the two receptors with
the two compounds. In any event, these results demon-
strate how a nuclear hormone binding assay may not
always be predictive of functional selectivity, particularly
when results are normalized across two different cell types
using a reference ligand.
Binding selectivity MCF-7 ERE Metallothionein-II
for ERb
ED50 (nM)
ED50 (nM)
17b-Estradiol 2
Example 6
0.00456ꢀ0.00092
n=2
>500
4.7ꢀ2.6
n=4
18
1
42
ꢁ58% efficacy
n=3
estradiol. Likewise, a comparison of apigenin 4 to its
ring constrained analogue 23 reveals a significant (>10-
fold) increase in binding affinity. The increase in binding
affinity that is observed maybe a consequence of con-
formational restriction and thus reduced entropy loss
upon binding. Alternatively, the enthalpy of binding
could be affected due to these compounds containing
different ring systems with different charge densities,
steric presentations, and so on. In any event, the
demonstration of significant binding affinity enhance-
ment over the parent compounds is noteworthy.
Constraining the ring systems of the naturally occurring
phytoestrogens apigenin and genistein through the intro-
duction of an oxa- or thia-bridge atom resulted in tetra-
cyclic compounds with increased ERb affinity and/or ERb
selectivity in a ligand binding assay. The phytoestrogen
obtained by oxa-constraint of genistein is the natural pro-
duct lupinalbin A. This compound proved to be a high
affinity ligand for ERb with an IC50 below that of 17b-
estradiol. The tetracyclic analogue of apigenin, compound
6, achieved ERb selectivity of over 40-fold in the ligand
binding assay and was subsequently tested in cellular sys-
tems responsive to either ERa or ERb. In the MCF-7 cell
expressing only ERa, compound 6 displayed an EC50 of
>500 nM. In SAOS-2 cells, compound 6 induced
metallothionein expression (an ERb specific response)
with an EC50=18 nM. This difference in potency between
ERa and ERb dependent functional response was much
more dramatic when normalized to 17b-estradiol and
indicate that compound 6 could be a useful tool for the
further elucidation of ERb pharmacology.
Replacement of the oxygen bridging atom with a sulfur
atom appears to increase binding affinity (compare com-
pounds 13 with 17 and 14 with 24). However, substitution
of the 3-position with either a bromo atom or cyano group
appeared to be very detrimental to good binding affinity
(compare compounds 20 with 15 and 25 with 6) (Table 1).
Functional selectivity of compound 6
Interestingly, all of the compounds for which binding
affinities could be determined showed at least some
selectivity for ERb. The most selective compound tested
was the tetracycle 6. It was more selective than its par-
ent, apigenin 4, and on a par with genistein 3 (the most
ERb selective phytoestrogen we have evaluated). Since 6
showed promising binding selectivity for ERb, we chose
to further evaluate it in our cell-based assays. In order
to measure potency and efficacy on ERb in a cellular
system, we looked for metallothionein induction in
SAOS-2 cells, a response we have shown is ERb specific.
For functional ERa evaluation we examined the ability
of selected compounds to stimulate reporter gene activity
in an MCF-7 ERE-tk luciferase assay since these cells
express ERa only and thus allow for a functional read-
out of ERa activity. In the metallothionein assay, com-
pound 6 is active on ERb with full efficacy and slightly
reduced potency relative to 17b-estradiol (Table 2).
However, in the ERa-MCF-7 cell line, compound 6 has
an EC50 >500 nM with approximately 60% efficacy. In
contrast, the non-selective steroid 17b-estradiol achieves
full efficacy in this cell system at 0.0045 nM. Thus in the
agonist mode, 17b-estradiol is at least 100,000 times
more potent than compound 6 on an ERa driven
response while it is only about 4 times as potent on the
ERb specific response in SAOS-2 cells (18 nM vs
4.7 nM). This indicates that despite its relatively modest
binding selectivity, compound 6 has great functional
selectivity across these two cell lines when normalized to
References and Notes
1. (a) Toran-Allerand, C. D. J. Steroid Biochem. Mol. Biol.
1996, 56, 169. (b) Kuiper, G. G. J. M.; Carlquist, M.; Gus-
tafsson, J.-A. Sci. Med. (Philadelphia) 1998, 5, 36. (c) Ritzen,
E. M.; Nilsson, O.; Holst, M.; Savendahl, L.; Wroblewski, J.
J. Steroid Biochem. Mol. Biol. 2000, 74, 383. (d) Abney, T. O.
Mol. Andol. 1992, 4, 281.
2. (a) Francoeur, R. T. Marriage Fam. Rev. 1993, 18, 135. (b)
Saunders, L. Fam. Plan. Resume 1977, 1, 18.
3. (a) Grese, T. A.; Dodge, J. A. Ann. Rep. Med. Chem.
1996, 31, 181. (b) Grese, T. A.; Dodge, J. A. Curr. Pharm.
Des. 1998, 4, 71. (c) Miller, C. P. Curr. Pharm. Des. 2002, 8,
2089.
4. (a) Doeren, M.; Schneider, H. P. G. Maturitas 1996, 25, 99.
(b) Tavani, A.; La Vecchia, C. Drugs Aging 1999, 14, 347.
5. Writing groupfor the Women’s Health Initiative JAMA
2002, 288, 321.
6. Sato, M.; Glasebrook, A. L.; Bryant, H. U. J. Bone Min.
Metab. 1994, 12, S9.
7. Clemett, D.; Spencer, C. M. Drugs 2000, 60, 370.
8. Kuiper, G. G.; Enmark, E.; Pelto-Huikko, M.; Nilsson, S.;
Gustafsson, J. A. Proc. Nat. Acad. Sci. U.S.A. 1996, 93, 5925.
9. Leygue, E.; Helmut, D.; Biao, L.; Glor, C.; Watson, P. H.;
Murphy, L. C. J. Clin. End. Metab. 1998, 83, 3754.
10. Pike, A. C. W.; Brzozowski, A. M.; Hubbard, R. E.; Pike,
A. C. W.; Brzozowski, A. M.; Hubbard, R. E.; Bonn, T.;
Thorsell, A.; Engstrom, O.; Ljunggren, J.; Gustafsson, J.;
Carlquist, M. EMBO J. 1999, 18, 4608.