3214
K. Ohta et al. / Bioorg. Med. Chem. Lett. 25 (2015) 3213–3216
H3C
O
N
CH3
n
N
O
N
O
n = 0 ~ 2
N
O
HO
4
O
N
ER partial agonist candidates
O
OH
H
HO
R
S
HO
Figure 3. Structures of novel ER partial agonist candidates 4.
R = H Tamoxifen
R = OH 4-Hydroxytamoxifen
Raloxifene
Bazedoxifene
in 94% yield; its biological activity was evaluated and compared
with that of bisphenol 12, an intermediate of 1 (Scheme 1).
Compound 4c was obtained by stepwise synthesis involving two
SN2 reactions of dibromobutane. One bromine atom was changed
to a phenolic hydroxyl group (7) in 62% yield, and the other was
reacted with dimethylamine to afford 11 in 78% yield.
Demethylation of 11 with BBr3 afforded 4c in 34% yield.
Figure 1. Structures of clinically used SERMs.
9,10-dimethyl-m-carborane cage is effective for obtaining ER par-
tial agonist activity with very low maximal efficacy, as well as
for increasing ER-binding affinity (compound 3, Fig. 2).15 This
may be due to a geometry change of the alkylamino group remo-
tely induced by steric repulsion between the bulky hydrophobic
structure and amino acid residues surrounding the ER ligand-bind-
ing domain (LBD). Although the 9,10-dimethyl-m-carborane cage
of 3 seems to be more promising as a hydrophobic structure for
ER partial agonist discovery than m-carborane, preparation of
1,7-diaryl-9,10-dimethyl-m-carborane derivatives is synthetically
difficult in that coupling reaction of 9,10-dimethyl-m-carborane
with aryl iodides affords the products in very low yield.
ER-binding affinity was evaluated by means of competitive
binding assay using human recombinant ER
a
and [6,7-3H]
17b-estradiol. Relative binding affinity (RBA) values of the test
compounds are summarized in Table 1.17 Compounds 1 and 12
showed low and high RBA values of 1.5 and 110, respectively,
which are close to the previously reported values of 1.1 and 106,
respectively.12 Although the binding affinity of the parent bisphe-
nol 9 is similar to that of 12, compound 4a with an N,N-dimethy-
laminoethoxy side chain showed 5 times more potent ERa-
Therefore, we selected the m-carborane cage as a hydrophobic
structure for the preparation of ER partial agonists and designed
derivatives 4 in which the alkylamino side chain of 1 is transferred
to a neighboring carbon (Fig. 3).
Scheme 1 summarizes the synthesis of m-carborane-containing
ER partial agonist candidates 4. m-Carborane 5 was treated with n-
BuLi, and then transformed into C-copper-p-carborane, which was
reacted with 4-iodoanisole in the presence of pyridine as a ligand
binding than the corresponding p-substituted derivative 1. Our
previous results showed that extension of the alkylamino chain
of 1 has little influence on RBA values, but the same modification
of the alkylamino chain of 4 led to a remarkable enhancement of
RBA value, and the RBA of 4c was 83.18 These results suggest that
an alkylamino side chain at the meta position fits well into the cav-
ity of the ER
a
LBD, and the terminal tertiary amino group of 4c
LBD.
forms hydrogen bonds with amino acid residues of the ER
a
of copper to afford 4-methoxyphenyl-m-carborane
6 in 71%
Next, the functional activities of the test compounds were eval-
uated by means of cell proliferation assays using MCF-7 cell lines
that show ER-dependent growth.17 Table 2 summarizes EC50 and
IC50 values as parameters of the agonist and antagonist activities
of the test compounds, respectively. Agonist activity is also shown
as relative maximal efficacy (Emax), based on estradiol as 100%.
Bisphenol 9 showed similar EC50 and Emax values to 12. Both
bisphenols showed no ER-antagonist activity and acted as ER full
agonists, not as side-chainless partial agonists like BE360. The lead
compound 1 showed moderate agonist activity and its Emax value
was 78%, which means it has lower maximal efficacy than E2.
Compound 1 antagonized MCF-7 cell proliferation induced by
yield.16 Next, coupling reaction of 6 with TBS-protected iodophenol
under the same conditions, followed by deprotection of the TBS
group, afforded key intermediate diaryl-m-carborane 7 in 71%
yield. Dimethylaminoethyl and dimethylaminopropyl groups were
introduced into 7 by using the corresponding alkyl halides in 33%
and 24% yields, respectively. Demethylation of 8 with BBr3 afforded
the desired compounds 4a and 4b in 45% and 74% yields, respec-
tively. Compound 7 was reacted with BBr3 to afford bisphenol 9
HO
OH
0.1 nM of E2 with an IC50 value of 4.4 lM. Compound 4a, which
has the same side chain as the lead compound 1, but at the meta
position, showed potent ER agonist activity (EC50 = 4.7 nM) and a
= C
= BH
unmarked
vertices
HO
HO
low Emax value of 63%. The IC50 value of 4a was 6.5 lM, which is
BE360
17β-estradiol (E2)
similar to that of 1. Compound 4b with a dimethylaminopropyl
group showed the lowest EC50 value and the lowest Emax value
among the tested compounds. In addition, compound 4b showed
10 times more potent ER antagonist activity than the lead com-
pound 1. Although dimethylaminobutyl derivative 4c showed the
ER partial agonist (in vitro)
SERM (in vivo)
CH3
N
CH3
N
R
H3C
O
O
H3C
O
greatest ERa-binding affinity, its biological activities parameters,
EC50, Emax, and IC50, are similar to those of compound 4a. These
results confirm that the dimethylaminopropyl group is the most
suitable ER partial agonist activity-inducing substituent in the
series of m-carborane-containing m-substituted derivatives 4.
The ER-antagonist activity of 4b was more potent than that of 3,
which contains the 9,12-dimethyl-m-carborane cage (IC50 of
CH3
CH3
HO
HO
HO
1
2
3
ER partial agonist
ER agonist
Potent ER partial agonist
3 = 0.88 lM). The low IC50 value of 4b suggested that the side chain
serves to inhibit binding of co-activators by moving helix-12 of ER
Figure 2. Structures of 17b-estradiol (E2) and m-carborane-containing ER
to an unfavorable position. However, compound 4b has a higher
modulators.