Preparation and evaluation of carborane analogues of tamoxifen

Article abstract of DOI:10.1021/jm100758j

A stereoselective synthesis of closo carborane analogues of tamoxifen was developed where the products represent a new approach to developing metabolically robust SERMs. The A-ring found in the backbone of tamoxifen was replaced with an ortho carborane cluster; the product was determined to be the desired Z isomer, which showed superior chemical stability to tamoxifen both in solution and in the solid state. By use of microwave heating, it was possible to convert some of the Z carborane tamoxifen analogue to the corresponding E isomer. Cell growth assays using both isomers and a carborane that is known to target the ER were conducted using estrogen receptor (ER) positive and ER negative human breast cancer cells with and without the presence of estradiol (E2). The Z carborane isomer was able to inhibit cell proliferation better than tamoxifen in an E2 free environment, while the E isomer inhibited cell growth better than tamoxifen when E2 was present.

Full text of DOI:10.1021/jm100758j

8012 J. Med. Chem. 2010, 53, 8012–8020
DOI: 10.1021/jm100758j
Preparation and Evaluation of Carborane Analogues of Tamoxifen
Michael L. Beer,† Jennifer Lemon,‡ and John F. Valliant*,†,‡
†Department of Chemistry and Chemical Biology and ‡Department of Medical Physics and Applied Radiation Sciences,
McMaster University, 1280 Main Street West, Hamilton, Ontario L8S 4M1, Canada
Received June 21, 2010
A stereoselective synthesis of closo carborane analogues of tamoxifen was developed where the products
represent a new approach to developing metabolically robust SERMs. The A-ring found in the backbone of
tamoxifen was replaced with an ortho carborane cluster; the product was determined to be the desired
Z isomer, which showed superior chemical stability to tamoxifen both in solution and in the solid state.
By use of microwave heating, it was possible to convert some of the Z carborane tamoxifen analogue to
the corresponding E isomer. Cell growth assays using both isomers and a carborane that is known to
target the ER were conducted using estrogen receptor (ER) positive and ER negative human breast
cancer cells with and without the presence of estradiol (E2). The Z carborane isomer was able to inhibit
cell proliferation better than tamoxifen in an E2 free environment, while the E isomer inhibited cell
growth better than tamoxifen when E2 was present.
Introduction
Tamoxifen (1, Figure 1) is currently used as a frontline treat-
ment in cases of hormone dependent breast cancer. The agent is
described by the World Health Organization (WHO) as an
essential drug for the treatment of breast cancer.1 In addition to
acute cancer therapy, tamoxifen is used as a preventative
treatment2 for high risk woman as well as in long-term adju-
vant therapy.3 The use of tamoxifen however has also been
associated with increased risk in developing endometrial4,5
Figure 1. Tamoxifen with aryl group labeling scheme.
and uterine6 cancers as well as increased possibility of suffer-
used previously to prepare boron neutron capture therapy
ing a pulmonary embolism, stroke, or deep vein thrombosis.7
(BNCT) agents as a result of their high boron content.17-20
In light of this, several tamoxifen analogues (Figure 2) have
What is less commonly known is that carboranes are interest-
been developed in hopes of retaining the benefits of the parent
ing hydrophobic structural motifs for use in medicinal chem-
drug while reducing possible side effects.
istry because of their unique 3D structures, ease of synthesis
Some of the early and more notable tamoxifen analogues
and functionalization, low toxicity, and resistance to catabo-
include raloxifene (2),8 idoxifene (3),9 and the tamoxifen metab-
lism in vivo.21
olite 4-hydroxytamoxifen (4).10 Many of the analogues re-
To date, there have been no reports of a closo carborane
tain key structural elements of tamoxifen (1) that are critical
analogue of tamoxifen. We previously reported the synthesis
in maintaining ER binding. This includes a pseudo-estradiol-
of a nido carborane analogue but were not able to prepare the
like structure containing a hydrophobic core and several
complementary closo structures or to evaluate the E versus Z
aromatic substituents around a central core.11,12 There re-
isomers.23 Given the promising results of simple carborane
mains an active search for robust tamoxifen analogues that are
phenol derivatives in binding to the ER, a synthetic method to
more effective selective estrogen receptor modulators
E and Z carborane analogues of tamoxifen (Scheme 1) was
(SERMsa) particularly for patients who are likely to be using
developed and the products were tested in ER positive and
the agents for prolonged periods of time.
negative cell lines.
Endo et al. have reported several arylcarborane derivatives
as estrogen analogues that have been shown to bind the estrogen
Results and Discussion
receptor13-16 (Figure 3) where the key structural components
needed for binding are a phenol ring and a closo carborane unit.
Carboranes are clusters of boron and carbon that have been
Synthesis and Characterization. Carboranes are readily
prepared from alkynes. Consequently a stereoselective meth-
od was developed to prepare the ene-yne 11 (Scheme 1). The
preparation of the methoxy ketone 7 was adapted from
Smyths and Corby24 and was successfully scaled to produce
multigram quantities (>100 g) where the product was
combined with TMS-protected lithium acetylide. The result-
ing tertiary alcohol 8 was isolated as pale yellow oil and used
*To whom correspondence should be addressed. Phone: 905-525-
9140, extension 20182. Fax: 905-522-7776. E-mail: valliant@mcmaster.
ca.
a Abbreviations: ER, estrogen receptor; E2, estradiol; SERM, selec-
tive estrogen receptor modulator; TMS, trimethylsilyl; BNCT, boron
neutron capture therapy; NOE, nuclear Overhauser effect.
r
pubs.acs.org/jmc
Published on Web 10/28/2010
2010 American Chemical Society

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22 8013
Figure 2. Tamoxifen analogues raloxifen (2), idoxifen (3), and 4-hydroxytamoxifen (4).
1.06 (Z) and 0.92 ppm (E) in Figure 4). A broad peak was
observed at 3.15 ppm that is attributable to the C-H of the
carborane cage, while the B-H groups of the carborane form
another broad signal spanning from 1 to 4 ppm.26 The peaks
in the 11B NMR spectrum and the MS isotope pattern were
consistent with the presence of a closo carborane cage, while
the IR spectrum showed B-H stretches at 2605 cm-1
.
X-ray quality single crystals of 13 were obtained by
dissolving the white solid in boiling petroleum ether, fol-
lowed by slow evaporation over 9 days. The crystal structure
of the closo carborane was consistent with the results of the
NOE experiments and confirmed that the product was in fact
the Z-isomer. The average C-C bond length of the carbor-
Figure 3. The closo carborane derivatives that bind the estrogen
receptor.22
without further purification. It was immediately treated with
thionyl chloride in pyridine at -78 °C and the corresponding
protected ene-yne was isolated in 89% yield. The 1H NMR of
9/10 showed a Z/E ratio of 15:1 which is consistent with the
elimination reaction occurring via a concerted E2 elimination
mechanism.
The TMS protecting group was removed by treating 9 and
10 with sodium methoxide in methanol (0.30 M). The reac-
tion mixture was stirred for 24 h whereupon analysis of the
mixture by thin layer chromatography (TLC) indicated
complete consumption of the starting material. The light
sensitive products 11 and 12 were isolated in 81% yield as a
mixture via simple liquid-liquid extraction (dichloromethane/
water). By use of 1H nuclear Overhauser effect (NOE)
spectroscopy, the major product was clearly identified as
the Z isomer. Integration of the signals arising from the
protons on the ethyl group (Figure 4) indicated that that the
Z/E ratio was 93:7. The residual E isomer was later removed
following isolation of the target carborane.
˚
ane was 1.600(2) A while the average B-B and C-B bond
˚
lengths were 1.782(3) and 1.719(2) A, respectively, which are
similar to the values reported for simple arylcarboranes.23
Removal of the methoxy protecting group in 13 was carried
out using BBr3 in ether to produce the phenol closo carborane 14
in 57% yield after recrystallization. The removal of the methoxy
protecting group proved to be difficult and required a longer
reaction time than those generally described in the literature.27
This may be a result of the extended conjugation in 13 which
would reduce the electron density on the methoxy oxygen,
thereby slowing the rate of reaction. Despite the poor reactivity,
the 1H NMR spectrum of 14 clearly indicated the formation of
the phenol through the loss of the signal associated with the
methoxy group and appearance of signal associated with the
phenol (δ 5.30 ppm). The IR spectrum further confirmed the
formation of the phenol group (ν=3573 cm-1).
Single crystals of 14 were obtained by slow evaporation of
a methanol solution, and the crystal structure (Figure 6)
confirmed that 14 was the Z-isomer. The data obtained from
the crystal structure allowed for comparison of the structural
backbone of tamoxifen and the closo carborane analogue. From
this information it was apparent that the introduction of the
carborane had negligible effect on the structure of the tamoxifen
backbone. The C10-C90 as well as the C10-C20 and C90-C100
bond lengths were found to be 1.348(2), 1.504(2), and 1.502(2)
With significant quantities in hand of the acetylene deri-
vative 11 having the appropriate stereochemistry, the alkyne
insertion reaction was performed to generate the desired
carborane. Initial attempts to obtain 13 utilizing a traditional
alkyne insertion reaction failed to yield the target compound.
This is likely the result of a combination of the highly conju-
gated nature of the ene-yne, which would reduce the overall
activity of the compound to carborane formation, and the
steric bulk around the reaction center. The ultimately suc-
cessful approach involved the use of 1-butyl-3-methylimida-
zolium chloride (Bmim-Cl), an ionic liquid.25 Decaborane
and catalytic amounts of Bmim-Cl were dissolved in toluene
to create a biphasic mixture to which 11 and 12 were added,
and the mixture was heated at reflux for 12 h under argon.
The removal of the reaction solvent yielded a dark solid which
was purified via silica gel chromatography, and recrystalliza-
tion gave the product as a white solid in 25% yield.
˚
A, respectively. The corresponding bond lengths in tamoxifen
28
˚
were found to be 1.347, 1.497, and 1.492 A.
When compounds 9-12 were dissolved in a solvent and
exposed to light, they isomerized and degraded to form a
mixture of compounds. In contrast, compound 14 could be
stored for several months as a solid in an open container
and as a methanol solution exposed to sunlight without any
evidence of degradation. To further investigate the stability
of 14, a temperature course study was conducted (Table 1)
by dissolving the compound in 95% ethanol and heating in a
1
The H NMR of 13 (Figure 5) was similar to that of the
deprotected acetylene derivative with the exception that purifi-
cation led to the isolation of a single Z isomer. This was evident
looking at the triplet at 0.92 ppm (in Figure 5 for 13) which
lacks the corresponding signal from the E isomer (as seen at
1
Biotage Initiator 60 microwave. The H NMR of the reac-
tion mixture showed that the compound did not degrade
but in fact isomerized to form a mixture of Z and E isomers

8014 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22
Beer et al.
Figure 4. 1H NMR of 11/12 following liquid-liquid extraction. Integration of the triplets in the highlighted region show a Z/E ratio of 15:1
(600 MHz, CDCl3).
Scheme 1. Synthesis of 14a
a (a) (1) TMS-acetylene, n-BuLi, -78°C, THF; (2) sat. ammonium chloride; (b) pyridine, thionyl chloride, -78 °C; (c) (1) NaOMe, MeOH; (2) H2O;
(d)1-butyl-3-methylimidazolium chloride, B10H14, toluene; (e) BBr3, dichloromethane, -78 °C.
(Figure 7). The relative amount of each isomer present was
determined by LC-MS.
In order to determine the identities of the two signals
observed in the HPLC of 14 (Figure 8) following heating, the

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22 8015
Figure 5. 1H NMR of 13 (600 MHz, CDCl3).
Figure 7. Thermal E-Z isomerization of 14 to give 15. The struc-
ture of molecule 16 is shown.
was the E-isomer 15 as opposed to the starting Z-isomer
(Figure 9). The most notable differences in the NMR were
the upfield shift of the signals associated with the methylene
group from 2.80 ppm in 14 to 1.89 ppm in 15 and the
significant (0.5 ppm) downfield shift of the aromatic protons
of 15 relative to those in 14.
Biological Evaluation. Prior to screening, the purities of 14,
15, and 1-(4-hydroxyphenyl)-1,2-dicarba-closo-dodecabor-
ane (16) were determined by reversed phase HPLC and
found to be 98%, 97%, and 99%, respectively. Their ability
to effect cell growth was evaluated using MCF-7 cells which
are known to overexpress the estrogen receptor (ER)29 as
well as the MDA MB 231 cells that are known to have low
endogenous ER expression.27 Assays were done in compar-
ison to 16, a compound that has been previously prepared in
our group23 and by Endo et al.30 which has shown the
ability to bind to ER. Compounds were evaluated at five
different concentrations (0.001, 0.01, 0.1, 1, and 10 μM) in
the presence and absence of estradiol (E2) at four time
points (1, 4, 7, and 10 days). Cell proliferation assays
using MDA MB 231 cells showed no appreciable difference
between the control samples and those incubated with 14,
15, or 16.
Figure 6. Crystal structure and atomic numbering scheme for 14
(50% thermal ellipsoids). Hydrogen atoms were omitted for clarity.
Table 1. E/Z Isomerization of 14 in 95% Ethanol Using Microwave
Heating
temp (°C)
time (min)
Z (14) (%)
E (15) (%)
80
100
120
140
160
180
20
20
20
20
20
20
100
100
100
99
0
0
0
1
85
56
15
44
peak eluting at 20.8 min was isolated and the product fully
characterized. The 1H NMR spectrum showed that the product

8016 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22
Beer et al.
Figure 8. HPLC chromatogram of 14 following microwave heating at 180 °C for 20 min in 95% ethanol. Elution conditions were as follows.
Solvent A = water, solvent B = acetonitrile. Gradient elution: 0-10 min, 80 f 20% A; 10-25 min, 20 f 0% A. Column: Zorbax RX-C18 (4.6
mm ꢀ 250 mm). Flow rate: 1 mL/min.
Figure 9. 1H NMR spectra of 14 (top) and 15 (bottom) (600 MHz, CDCl3).
In the MCF-7 cells, at 1 μM, 14 without E2 present
(Figure 10A) showed nearly identical results compared to
tamoxifen on all days, with day 10 generating a percent of
control of 7.0 ( 0.2% and 5.8 ( 0.6%, respectively

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22 8017
Figure 10. The results of cell proliferation assays reported a percentage of control for compounds 1, 14, 15, and 16 at days 1, 4, 7, and 10: (A)
1 μM, (B) 0.001 μM, (C) 1 μM with E2, (D) 0.001 μM with E2. Parts A and B use growth data with no additives, and parts C and D use growth
data plus 10 nM E2. Refer to Supporting Information for 0.01 and 0.1 μM graphs.
(p < 0.15). At 0.001 μM, tamoxifen performed significantly
better than 14, showing on day 10 an average of 49.2 (
2.10% of control cells compared to 60.7 ( 0.9% for 14
(Figure 10B). In the presence of 10 nM E2 (Figure 10C) the
effect of 1.0 μM tamoxifen changed significantly over time
with percent of control ranging from 101.7 ( 11.7% on day
1 to 13.3 ( 0.4% (p < 0.006) on day 10, which differs
considerably from that for 14 (29.2 ( 4.4% to 17.1 ( 0.4%;
p < 0.00002) under the same conditions. This result indicates
that the onset of cell growth inhibition by 14 is more rapid
than tamoxifen, which may indicate an inhibitory ER me-
chanism and/or a cytotoxic response. From Figure 10C it is
evident that there is a slight difference in cell proliferation
between tamoxifen (1) and 14 (13.3 ( 0.4 vs 17.1 ( 0.4, p <
0.005) by day 10, which, although statistically different, is
unlikely to be of biological significance. The plot also
illustrates that 15 exhibits inhibitory action similar to tamox-
ifen, in that it is most effective at later time points, and its
ability to inhibit cell growth was superior to 14 after day 4.
The E tamoxifen analogue (15) exhibits superior inhibition
of cellular proliferation compared to tamoxifen on days 4, 7,
and 10, showing an average percent of control of 20.5 (
2.6%, 11.9 ( 0.4%, and 8.0 ( 0.9% in comparison to 30.6 (
1.4% (p < 0.04), 20.7 ( 0.8% (p < 0.002), and 13.3 ( 0.4%
(p < 0.02), respectively. At 1 μM 15 with 10 nM E2, 15 shows
greater inhibition of cell proliferation compared to 1, 14, and
16; however, as the concentration of each drug was decreased
with the level of E2 remaining constant, relative cell prolif-
eration increased. At 0.001 μM compound þ10 nM E2
(Figure 10D), at day 10, 16 showed the fewest number
of viable cells (42.3 ( 1.0%, p < 0.000 07) compared to
tamoxifen (66.4 ( 0.9%), followed by 15 (53.8 ( 0.7%,
p < 0.0004) and 14 (67.1 ( 1.2%, p < 0.7), which did not
differ significantly from tamoxifen. In the presence of 10 nM
E2, 14 and 16 show rapid reduction in cell numbers from day
1, which may indicate a cytotoxic response instead of solely
an ER-based response. The proliferation values at 0.01 and
0.1 μM are between 1 and 0.001 μM, with similar patterns of
cell growth inhibition over the 10-day incubation period.
Graphs of cell proliferation for 0.01 and 0.1 μM concentra-
tions can be found in Supporting Information.
Although there were significant differences between the
ER-positive and -negative expressing cell lines in their re-
sponse to the carborane compounds, it cannot be conclu-
sively stated that ER-dependent pathways were the only
mechanism of action of these analogues. It is well document-
ed that tamoxifen has ER-independent mechanisms, which
can reduce cell numbers, primarily through the induction of
apoptotic pathways.31-35 Now that it has been established
that the carborane derivatives exhibit comparable or super-
ior cytotoxic responses, a panel of receptor binding assays
including ER binding measurements are currently underway
to better elucidate the mechanism of action of the boron
compounds.
Conclusion
The stereoselective synthesis of a closo carborane tamoxifen
analogue was achieved in reasonable overall yield. It was
determined that isomerization could be carried out using
microwave heating to obtain a 50:50 mixture of the E and Z
closo carborane isomers. Despite the isomerization, the closo
carborane tamoxifen analogue has greater stability toward
degradation than tamoxifen itself. The results of the cell
proliferation assay indicate that the Z carborane tamoxifen
analogue (14) shows similar inhibition compared to tamoxifen

8018 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22
Beer et al.
in an E2 free environment, while in the presence of E2 the E
carborane tamoxifen analogue (15) showed the lowest num-
ber of viable cells remaining after 4 days. These data in
combination warrant further evaluation of the carborane
analogues of tamoxifen.
Crystallographic Data Centre via www.ccdc.cam.ac.uk/da-
ta_request/cif.
Syntheses. (Z/E)-(3-(4-Methoxyphenyl)-4-phenylhex-3-en-
1-ynyl)trimethylsilane (9/10). Trimethylsilylacetylene (35 mL,
250 mmol) was added to dry THF (750 mL) that had been
chilled to -78 °C. To this, n-BuLi (100 mL, 0.25 mol) was
added dropwise to form a clear colorless solution. Compound
7 (63.5 g, 250 mmol) dissolved in dry THF (500 mL) was added
over 30 min whereupon the mixture turned bright pink. The
reaction mixture was allowed to warm to room temperature
over 12 h. To the pale yellow solution chilled to 0 °C,
saturated ammonium chloride solution (400 mL) was added
followed by water (800 mL). The mixture was separated into
two fractions, and each was extracted with diethyl ether (3 ꢀ
500 mL) which was pooled and dried over anhydrous sodium
sulfate and evaporated giving a yellow oil. The oil (6) was
dissolved in 40 mL of anhydrous pyridine and THF (500 mL).
The solution was cooled to -78 °C, and thionyl chloride
(16 mL, 185 mmol) was added dropwise over 30 min. The
reaction mixture was allowed to warm to room temperature
over 12 h. Water (200 mL) was added to the pale yellow
solution, followed by 2 M HCl (100 mL). The organic and
aqueous layers were separated, and the organic fraction was
washed with 2.0 M HCl (2 ꢀ 600 mL). The aqueous fractions
were washed with diethyl ether (3 ꢀ 500 mL) and all organic
fractions combined, dried over anhydrous sodium sulfate,
and evaporated to give a brown oil (75 g, 89%). TLC (7.5:1
hexanes/ethyl acetate) Rf=0.78. 1H NMR (200 MHz, CDCl3):
δ 7.25-6.60 (m, 9H, aryl), 3.75 (s, 3H, -O-CH3), 2.90 (q, J =
7.5 Hz, 2H, CH3), 1.05 (t, J = 7.5 Hz, 3H, CH2), 0.25 (s, 9H,
-S-(CH3)3). 13C NMR (50 MHz, CDCl3): δ 158.1, 151.8,
140.8, 131.0, 129.0, 128.0, 126.8, 119.4, 113.0, 105.9, 98.3,
55.1, 31.4, 0.1. m/z calculated for [M þ H]þ C22H27OSi:
335.1831. HRMS EIþ [M þ H]þ 335.1847. FTIR (NaCl,
cm-1) ν: 1608, 2138, 2962.
(Z/E)-1-Methoxy-4-(4-phenylhex-3-en-1-yn-3-yl)benzene (11/12).
Compounds 9 and 10 (75 g, 224.5 mmol) were dissolved in a
solution of sodium methoxide (16 g, 305.5 mmol) in methanol
(1 L). The mixture was stirred for 24 h at room temperature.
Water (500 mL) was added, and the mixture was extracted with
dichloromethane (3 ꢀ 600 mL), which in turn was extracted with
water (3 ꢀ 600 mL). The organic layer was evaporated to give a
yellow oil. Yield (53 g, 81%). TLC (7.5:1 hexanes/ethyl acetate):
Rf = 0.46. 1H NMR (600 MHz, CDCl3): δ 7.21-6.65 (m, 9H,
aryl), 3.72 (s, 3H, OCH3), 3.31 (s, 1H, CH), 2.94 (q, J = 7.5 Hz,
2H, CH2), 1.06 (t, J = 7.5 Hz, 3H, CH3). 13C NMR (150 MHz,
CDCl3): δ158.3, 152.4, 140.6, 131.0, 130.0, 128.6, 128.1, 118.5,
113.2, 84.5, 81.4, 55.2, 31.3, 12.6. m/z calculated for [M þ H]þ
C19H19O: 263.1436. HRMS [M þ H]þ EIþ 263.1430. FTIR
(NaCl, cm-1) ν: 1608, 2088, 2969, 3288.
Experimental Section
General. All reactions were carried out under argon that had
been passed through Drierite, using commercial grade solvents
dried using a Pure Solv MD-6 solvent purification system
(Innovative Technology Inc.). Chemicals were purchased from
Sigma-Aldrich or SGF chemicals and used without further
purification. Decaborane was purchased from Katchem. Reac-
tions requiring microwave heating were performed using a
Biotage Initiator 60 instrument. Compound 7 was prepared
following the literature procedure.24 1H, 13C, and 11B NMR
spectra were recorded on a Bruker AV600, Bruker DRX500, or
Bruker AV200 spectrometer with probe temperatures of 30, 25,
or 25 °C, respectively. 1H NMR chemical shifts are reported in
ppm relative to the residual proton signal of the NMR solvent.
Coupling constants (J) are reported in hertz (Hz). 13C chemical
shifts are reported in ppm relative to the carbon signal of the
solvent, while 11B chemical shifts are reported in ppm relative to
an external standard of BF3 Et2O. Thin layer chromatography
3
plates (Merck F254 silica gel on aluminum plates) were visual-
ized using 0.1% PdCl2 in 3 M HCl(aq) and/or UV light. Infrared
spectra were acquired using a BioRad FTS-40 FT-IR or Nicolet
510 FTIR spectrometer. High resolution mass spectra were
obtained on a Waters/micromass Q-ToF Ultima Global spec-
trometer. Low resolution LCMS were obtained on a Waters
2695 LC with a Quattro Ultima triple quadrupole mass spectro-
meter. Analytical and semipreparative HPLC was performed
using a Varian Pro Star model 330 PDA detector, model 410
autosampler, model 230 solvent delivery system, and model 710
fraction collector. Analysis was conducted using Agilent Zorbax
SB-C18 (2a, 4.6 mm ꢀ 250 mm, and 2b, 9.8 mm ꢀ 250 mm)
columns. Analytical HPLC experiments were monitored at
254 nm using a flow rate of 1.0 mL min-1 at 30 °C. Semipre-
parative HPLC experiments were monitored at 254 nm using a
flow rate of 4.0 mL min-1 at room temperature. The elution
protocol used was as follows. HPLC method A: solvent A =
water, solvent B = acetonitrile. Gradient elution: 0-10 min,
80% f 20% A; 10-25 min, 20% f 0% A. Compounds were
purified to at least 95% as determined by analytical HPLC.
Crystallographic Details. X-ray crystallographic data for 13
were collected from a single crystal frozen Paratone oil. Data
were collected at 173 K on a Bruker Smart36 Apex2Mo diffract-
ometer using an Apex2 detector (512 ꢀ 512 mode) at 5 cm
from the sample, three-circle D8 goniometer, 50 kV/30 mA fine
focused sealed tube with graphite monochromator, and
0.50 mm pinhole collimator. Data were collected with ω and φ
scans. Data for 14 was collected on a three-circle D8 Bruker
diffractometer equipped with a Bruker SMART 6000 CCD area
detector and a rotating anode utilizing cross-coupled parallel
focusing mirrors to provide monochromated Cu KR radiation
(Z)-1-(1,2-Dicarba-closo-dodecaborane-1-yl)-1-(4-methoxy-
phenyl)-2-phenylbut-1-ene (13). 1-Butyl-3-methylimidazolium
chloride (2.8 g, 89.0 mmol) and decaborane (12 g, 99 mmol)
were dissolved in dry toluene (500 mL), forming a biphasic
mixture. The solution containing a mixture of 11 and 12
(22.0 g, 84.0 mmol) dissolved in dry toluene was added
dropwise and the reaction mixture heated to reflux for 12 h.
The mixture was filtered and then subsequently evaporated to
yield a dark solid which was purified by column chromato-
graphy (100% hexanes f 7.5:1 hexane/ethyl acetate) to yield a
pale yellow liquid, which upon cooling yielded a white solid.
The white solid was recrystallized from low boiling petroleum
ether to yield white crystals (8.0 g, 25%). Mp: 120-121 °C.
˚
(λ = 1.541 78 A). Data processing for both compounds was
carried out by use of the program SAINT,37 while the program
SADABS38 was utilized for the scaling of diffraction data, the
application of a decay correction, and an empirical absorption
correction based on redundant reflections. The structures were
solved by using the direct-methods procedure in the Bruker
SHELXTL program library39 and refined by full-matrix least-
squares methods on F2. All non-hydrogen atoms were refined
using anisotropic thermal parameters, and hydrogen atoms
were located using the difference map and refined using iso-
tropic thermal parameters. CCDC 781744 and 781745 contain
the supplementary crystallographic data for this paper. These
data can be obtained free of charge from The Cambridge
1
TLC (7.5:1 hexanes/ethyl acetate): Rf = 0.55. H NMR (500
MHz, CDCl3): δ 7.2-6.6 (m, 9H, aryl), 3.67 (s, 3H, -O-CH3),
3.15 (s, 1H, Ccarborane-H), 2.95 (q, J = 7.0 Hz, 2H, CH2), 0.92
(t, J = 7.0 Hz, 3H, CH3). 11B{1H} NMR (160 MHz, CDCl3):
δ -3.7, -8.9, -10.1, -12.3, -14.2. 13C NMR (125 MHz,
CDCl3): δ 158.7, 151.2, 142.7, 131.8, 131.4, 129.4, 129.3, 128.5,
127.5, 126.1, 113.7, 63.2, 55.2, 27.7, 12.8. m/z calculated for

Article
Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22 8019
C19H28B10O: [M þ formate]- and [M þ acetate]- 425.3129
and 439.3286. HRMS TOF EI- [M þ formate]- and [M þ
acetate]- 425.3120 and 439.3281. FTIR (NaCl, cm-1): ν 1605,
2605, 3071.
surface was flushed 3 times to ensure the removal of all cells. An
aliquot (50 μL) of the resulting cell suspension was placed in a
microcentrifuge tube for counting. An equivalent volume of the
viability dye Trypan Blue was added to the cell suspension and
mixed well, and 10 μL of the resulting solution was pipetted onto
a hemocytometer. The cells within four grids were counted and
averaged to obtain a count of the total cells in each well. A Leitz
Dialux light microscope was used for the analysis.
(Z)-1-(1,2-Dicarba-closo-dodecaborane-1-yl)-1-(4-phenol)-2-phe-
nylbut-1-ene (14). Compound 13 (1.0 g, 2.63 mmol) was dis-
solved in dry dichloromethane and cooled to -78 °C. BBr3
(4.0 mL, 4.0 mmol) was added over 5 min, and the mixture was
allowed to warm to room temperature over 12 h and stirred for
an additional 36 h. The mixture was evaporated to dryness and
the residue redissolved in methanol and evaporated. The brown
solid was recrystallized from low boiling petroleum ether,
yielding 13 as a colorless solid (0.5 g, 57%). Mp: 179-181 °C.
Acknowledgment. This work was supported by funding
from the Natural Sciences and Engineering Research Council
(NSERC) of Canada and the Canadian Institutes for Health
Research (CIHR). We thank the staff of the X-ray Diffraction
Facility at McMaster University for performing the X-ray
analyses and preparing the structures for publication. The
authors also thank Dr. Laura Harrington for her assistance
with X-ray data and preparation of the manuscript.
1
TLC (7.5:1 hexanes/ethyl acetate): Rf = 0.27. H NMR (500
MHz, CDCl3): δ 6.9 -6.4 (m, 9H, aryl), 4.4 (s, 1H, -OH), 3.05 (s,
1H, Ccarborane-H), 2.80 (q, J = 7.5 Hz, 2H, CH2), 0.80 (t, J = 7.5
Hz, 3H, CH3). 13C NMR (125 MHz, CDCl3) δ 154.2, 150.5,
141.9, 131.2, 130.7, 128.2, 127.7, 126.8, 125.4, 114.5, 67.4, 62.4,
26.9, 25.0, 12.0. 11B{1H} NMR (160 MHz, C3D6O): δ -3.1,
-8.4, -9.2, -11.1, -13.2. m/z calculated for [M - H]- C18H25-
B10O-: 365.2918. HRMS TOF EI- [M - H]- 365.2910. FTIR
(NaCl, cm-1) ν: 1609, 2608, 3573.
(E)-1-(1,2-Dicarba-closo-dodecaborane-1-yl)-1-(4-phenol)-
2-phenylbut-1-ene (15). Compound 14 (72 mg, 0.196 mmol) was
dissolved in 95% ethanol (2.0 mL) and heated using microwave
irradiation at 185 °C for 25 min. An aliquot of the reaction
mixture (24 mg) was purified via semipreparative HPLC
(method A, column 2b), yielding a white solid (8.4 mg, 35%).
Mp: 144-145 °C. TLC (7.5:1 hexanes/ethyl acetate): Rf = 0.34.
1H NMR (600 MHz, CDCl3): δ 7.41 -6.85 (m, 9H, aryl), 4.86 (s,
1H, OH), 3.12 (s, 1H, carborane CH), 1.89 (q, J = 7.5 Hz,
2H, CH2), 0.69 (t, J = 7.5 Hz, 3H, CH3). 13C NMR (150
MHz, CDCl3): δ 155.2, 147.7, 138.8, 132.1, 131.5, 129.7,
128.4, 128.0, 127.7, 115.2, 59.1, 33.7, 11.7. 11B{1H} NMR (192
MHz, CDCl3): -3.5, -10.2, -11.1, -13.9. m/z calculated for
[M - H]- C18H25B10O- 365.2918. HRMS TOF EI- [M - H]-
365.2920.
Cell Lines and Tissue Culture. MCF-7 and MDA MB 231
human breast adenocarcinoma cells lines were obtained from
ATCC (Manassas, VA). Both cell lines were cultured in DMEM
without phenol red (CA12001-630, VWR International, Mis-
sissauga, Ontario, Canada) supplemented with 10% charcoal-
stripped fetal bovine serum (CA95039-622, VWR International),
1% L-glutamine (25030-081, Invitrogen, Mississauga, Ontario,
Canada), and 1% antibiotic/antimycotic (AB/AM) (15240-062,
Invitrogen). All cells were maintained at 37 °C in 5% CO2.
Growth Inhibition Study. MCF-7 and MDA MB 231 cell lines
were used to assess whether 14, 15, and 16 were effective
inhibitors of cell proliferation in ER positive and ER negative
expressing cells. Both cell lines were plated (5.0 ꢀ 104 cells) in
each well of six-well plates (9.51 cm2 growth surface) in 2 mL of
medium 24 h prior to the start of the experiment to allow for cell
adhesion to growth surface and repopulation of cell surface
receptors. Each cell line had 46 treatment groups (three replicate
wells per group) which included control (no treatment), 0.001,
0.01, 0.1, 1, and 10 μM estradiol, tamoxifen, 14, 15 or 16 alone,
or compound plus 10 nM estradiol (excluding estradiol group).
At the start of the experiment (day 0), the media were removed
from each culture and replaced with 2 mL of supplemented
media containing the appropriate concentration of the experi-
mental compound. The compound levels were maintained by
replacing the culture media with fresh compound containing
media every 3 days.
Supporting Information Available: Two crystallographic files
in CIF format; NMR, IR, and MS spectral data for all novel
compounds; and X-ray data for compounds 13 and 14. This
material is available free of charge via the Internet at http://
pubs.acs.org.
References
(1) Jordan, V. C. Tamoxifen: a most unlikely pioneering medicine.
Nat. Drug Rev. 2003, 2, 205–213.
(2) Fisher, B.; Costantino, J. P.; Wickerham, D. L.; Redmond,
C. K.; Kavanah, M.; Cronin, W. M.; Vogel, V.; Robidoux, A.;
Dimitrov, N.; Atkins, J.; Daly, M.; Wieand, S.; Tan-Chiu, E.;
Ford, L.; Wolmark, N. Tamoxifen for prevention of breast
cancer: report of the National Surgical Adjuvant Breast and
Bowel Project P-1 Study. J. Natl. Cancer Inst. 1998, 90 (18),
1371–1388.
(3) Gelber, R. D.; Cole, B. F.; Goldhirch, A.; Rose, C.; Fisher,
B.; Osborne, C. K.; Boccardo, F.; Gray, R.; Gordon, N. H.;
Bengtsson, N.; Sevelda, P. Adjuvant chemotherapy plus tamo-
xifen compared with tamoxifen alone for postmenopausal breast
cancer: meta-analysis of quality adjusted. Lancet 1996, 347, 1066–
1071.
(4) Bland, A. E.; Calingaert, B.; Secord, A. A.; Lee, P. S.; Valea, F. A.;
Berchuck, A.; Soper, J. T.; Havrilesky, L. Relationship between
tamoxifen use and high risk endometrial cancer histologic types.
Gynecol. Oncol. 2009, 112, 150–154.
(5) Yuvaraj, S.; Premkumar, V. G.; Shanthi, P.; Vijayasarathy, K.;
Gangadaran, S. G. D.; Sachdanandam, P. Effect of coenzyme Q10,
riboflavin and niacin on tamoxifen treated postmenopausal breast
cancer women with special reference to blood chemistry profiles.
Breast Cancer Res. Treat. 2009, 114, 377–384.
(6) Hoogendoorn, W. E.; Hollema, H.; van Boven, H. H.; Bergman,
E.; de Leeuw-Mantel, G.; Platteel, I.; Fles, R.; Nederlof, P. M.;
Mourits, M. J. E.; van Leeuwen, F. E.; Comprehensive Cancer
Centers TAMARISK-Group.. Prognosis of uterine corpus cancer
after tamoxifen treatment for breast cancer. Breast Cancer Res.
Treat. 2008, 112, 99–108.
(7) Dobrydneva, Y.; Weatherman, R. V.; Trebley, J. P.; Morrell,
M. M.; Fitzgerald, M. C.; Fichandler, C. E.; Chatterjie, N.;
Blackmore, P. F. Tamoxifen stimulates calcium entry into hu-
man platelets. J. Cardiovasc. Pharmacol. 2007, 50, 380–390.
(8) Draper, M. W.; Flowers, D. E.; Neild, J. A.; Huster, W. J.; Zerbe,
R. L. Antiestrogenic properties of raloxifene. Pharmacology 1995,
50, 209–17.
(9) Pace, P.; Jarman, M.; Phillips, D.; Hewer, A.; Bliss, J.; Coombes,
R. C. Idoxifene is equipotent to tamoxifen in inhibiting mammary
carcinogenesis but forms lower levels of hepatic DNA adducts. Br.
J. Cancer 1997, 76, 700–704.
(10) Borgna, J. L.; Rochefort, H. In vivo occupation of estrogen
receptors by hydroxylated metabolites of tamoxifen. Compt. Rend.
1979, 289, 1141–4.
Triplicate wells of each treatment group were harvested on
days 1, 4, 7, and 10 to determine total cell number/well. Media
were removed from each well and discarded, and the cells were
gently rinsed with 1 mL phosphate buffered solution (PBS).
Cells were released from the growth surface by the addition of
500 μL of 0.05% trypsin/EDTA to each well, followed by the
addition of 1 mL of media to inactivate the trypsin. The growth
(11) Wakeling, A. E.; Seater, S. R. Cancer Treat. Rep. 1980, 64, 741–744.
(12) Jordan, V. C. Metabolites of tamoxifen in animals and man:
identification, pharmacology, and significance. Breast Cancer
Res. Treat. 1982, 2, 123–138.
(13) Endo, Y.; Iijima, T.; Yamakoshi, Y.; Kubo, A.; Itai, A. Structure-
activity study of estrogenic agonists bearing dicarba-closo-dode-
caborane. Effect of geometry and separation distance of hydroxyl

8020 Journal of Medicinal Chemistry, 2010, Vol. 53, No. 22
Beer et al.
groups at the ends of molecules. Bioorg. Med. Chem. Lett. 1999, 9,
3313–3318.
(14) Endo, Y.; Iijima, T.; Yamakoshi, Y.; Yamaguchi, M.; Fukasawa, H.;
ShudoK. Potent estrogenic agonists bearing dicarba-closo-dode-
caborane as a hydrophobic pharmacophore. J. Med. Chem. 1999,
42, 1501–1504.
(26) Hegstrom, R. A.; Newton, M. D.; Potenza, J. A.; Lipscomb, W. N.
Chemical shifts of boron-11 in icosahedral carboranes. J. Am.
Chem. Soc. 1966, 88, 5340–5342.
(27) Rochefort, H.; Glondu, M.; Sahla, M. E.; Platet, N.; Garcia, M.
How to target estrogen receptor-negative breast cancer? Endocr.-
Relat. Cancer 2003, 10, 261–266.
(15) Endo, Y.; Iijima, Y.; Yamakoshi, Y.; Fukasawa, H.; Miyaura, C.;
Inada, M.; Kubo, A.; Itai, A. Potent estrogen agonists based on
carborane as a hydrophobic skeletal structure. A new medicinal
application of boron clusters. Chem. Biol. 2001, 8, 341–355.
(16) Endo, Y.; Yoshimi, T.; Miyaura, C. Boron clusters for medicinal
drug design: selective estrogen receptor modulators bearing car-
borane. Pure Appl. Chem. 2003, 75, 1197–1205.
(17) Tolpin, E. I.; Wong, H. S.; Lipscomb, W. K. Binding studies of
boron hydride derivatives to proteins for neutron capture therapy.
J. Med. Chem. 1974, 17, 792–796.
(28) Weeks, C. M.; Griffin, J. F.; Duax, W. L. Am. Crystallogr. Assoc.,
Ser. 2 1977, 5, 69.
(29) Dickson, R. B.; Bates, S. E.; McManaway, M. E.; Lippman, M. E.
Characterization of estrogen responsive transforming activity in
human breast cancer cell lines. Cancer Res. 1986, 46, 1707–1713.
(30) Endo, Y.; Yamamoto, K.; Kagechika, H. Utility of boron clusters
for drug design. Relation between estrogen receptor binding
affinity and hydrophobicity of phenols bearing various types of
carboranyl groups. Bioorg. Med. Chem. Lett. 2003, 13, 4089–
4092.
(18) Kim, Y. S.; Kane, R. R.; Beno, C. L.; Romano, S.; Mendez, G.;
Hawthorne, M. F. Synthesis of new building blocks for boron-rich
oligomers in boron neutron capture therapy (BNCT). II. Mono-
mers derived from 2,2-disubstituted-1,3-diols. Tetrahedron Lett.
1995, 36, 5147–5150.
(19) Meo, C. D.; Panza, L.; Campo, F.; Capitani, D.; Mannina, L.;
Banzato, A.; Rondina, M.; Rosato, A.; Crescenzi, V. Novel types
of carborane-carrier hyaluronan derivatives via click chemistry.
Macromol. Biosci. 2008, 8, 670–681.
(20) Wong, H. S.; Tolpin, E. I.; Lipscomb, W. N. Boron hydride
derivatives for neutron capture therapy. Antibody approach. J.
Med. Chem. 1974, 17, 785–791.
(21) Valliant, J. F.; Guenther, K. J.; King, A. S.; Morel, P.; Schaffer, P.;
Sogbein, O. O.; Stephenson, K. A. The medicinal chemistry of
carboranes. Coord. Chem. Rev. 2002, 232, 173–230.
(22) Endo, Y.; Yoshimi, T.; Miyaura, C. Boron clusters for medicinal
drug design: selective estrogen receptor modulators bearing car-
borane. Pure Appl. Chem. 2003, 75, 1197–1205.
(23) Causey, P. W.; Besanger, T. R.; Valliant, J. F. Synthesis and
screening of mono- and di-aryl technetium and rhenium metallo-
carboranes. A new class of probes for the estrogen receptor. J. Med.
Chem. 2008, 51, 2833–2844.
(31) Zheng, A.; Kallio, A.; Harkonen, P. Tamoxifen-induced rapid
death of MCF-7 breast cancer cells is mediated via extracellularly
signal-regulated kinase signaling and can be abrogated by estrogen.
Endocrine 2007, 148, 2764–2777.
(32) Zhang, W.; Couldwell, W. T.; Song, H.; Takano, T.; Lin, J. H. C.;
Nedergaard, M. Tamoxifen-induced enhancement of calcium sig-
naling in glioma and MCF-7 breast cancer cells. Cancer Res. 2000,
60, 5395–5400.
(33) Ferlini, C.; Scambia, M.; Marone, M.; Distefano, M.; Gaggini, C.;
Ferrandina, G.; Fattorossi, A.; Isola, G.; Benedetti Panici, P.;
Mancuso, S. Tamoxifen induces oxidative stress and apoptosis in
oestrogen receptor-negative human cancer cell lines. Br. J. Cancer
1999, 79, 257–263.
(34) Taylor, I. W.;Hodson, P. J.;Green, M. D.;Sutherland, R. L. Effectsof
tamoxifen on cell cycle progression of synchronous MCF-7 human
mammary carcinoma cells. Cancer Res. 1983, 43, 4007–4010.
(35) Lehenkari, P.; Parikka, V.; Rautiala, T. J.; Weckstrom, M.;
Dahllund, J.; Harkonen, P. L.; Vaananen, H. K. The effects of
tamoxifen and toremifene on bone cells involve changes in plasma
membrane ion conductance. J. Bone Miner. Res. 2003, 18, 473–481.
(36) SMART: Area-Detector Software; Siemens Industrial Automation,
Inc.: Madison, WI, 1993.
(24) Smyth, T. P.; Corby, B. W. Toward a clean alternative to Friedel-
Crafts acylation: in situ formation, observation, and reaction of an
acyl bis(trifluoroacetyl)phosphate and related structures. J. Org.
Chem. 1998, 63, 8946–8951.
(25) Kursari, U.; Li, Y.; Bradley, M. G.; Sneddon, L. G. Polyborane
reactions in ionic liquids: new efficient routes to functionalized
decaborane and o-carborane clusters. J. Am. Chem. Soc. 2004, 126,
8662–8663.
(37) SAINT: Area-Detector Integration; Siemens Industrial Automation,
Inc.: Madison, WI, 1995.
(38) SADABS: Area-Detector Absorption; Siemens Industrial Automa-
tion, Inc.: Madison, WI, 1996.
(39) Sheldrick, G. M. SHELXTL [Includes SHELXS97, SHELXL97,
CIFTAB ] Programs for Crystal Structure Analysis; Instit€ut f€ur
€
Anorganische Chemie der Universitat: Tammanstrasse 4, D-3400
€
Gottingen, Germany, 1998.