Development of small-molecule cyclin D1-ablative agents

Article abstract of DOI:10.1021/jm060057h

Previously, we demonstrated that the peroxisome proliferator-activated receptor γ (PPARγ) agonist troglitazone mediated the repression of cyclin D1 in MCF-7 breast cancer cells by facilitating proteasome-facilitated proteolysis. This PPARγ-independent mec

Full text of DOI:10.1021/jm060057h

4684
J. Med. Chem. 2006, 49, 4684-4689
Development of Small-Molecule Cyclin D1-Ablative Agents
Jui-Wen Huang,† Chung-Wai Shiau,† Jian Yang,† Da-Sheng Wang,† Hao-Chieh Chiu,† Ching-Yu Chen,‡ and Ching-Shih Chen*,†
DiVision of Medicinal Chemistry and Pharmacognosy, College of Pharmacy, 336 Parks Hall, The Ohio State UniVersity, 500 West 12th
AVenue, Columbus, Ohio 43210, Department of Family Medicine, College of Medicine, National Taiwan UniVersity, Taipei, Taiwan, and The
DiVision of Gerontology Research, National Health Research Institutes, Taipei, Taiwan
ReceiVed January 17, 2006
Previously, we demonstrated that the peroxisome proliferator-activated receptor γ (PPARγ) agonist
troglitazone mediated the repression of cyclin D1 in MCF-7 breast cancer cells by facilitating proteasome-
facilitated proteolysis. This PPARγ-independent mechanism provided a molecular basis for using troglitazone
as scaffold to develop a novel class of cyclin D1-ablative agents. The proof of principle of this premise is
provided by ∆2TG, in which the introduction of a double bond adjacent to the thiazolidinedione ring abrogated
the PPARγ activity while retaining the activity in cyclin D1 repression. Structural optimization of ∆2TG
led to STG28 [(S)-5-(4-{[6-(allyloxy)-2,5,7,8-tetramethylchroman-2-yl]methoxy}-3-methoxybenzylidene)-
thiazolidine-2,4-dione], which exhibited low micromolar potency in ablating cyclin D1 and inhibiting MCF-7
cell proliferation. It is noteworthy that STG28 mediated the proteasomal degradation of cyclin D1 with a
high degree of specificity. Exposure to STG28 did not cause any appreciable change in the expression
levels of a series of other cyclins and CDK-dependent kinases. In light of the pivotal role of cyclin D1 in
promoting tumorigenesis and drug resistance, this novel cyclin D1-ablating agent may have therapeutic
relevance in cancer therapy.
Introduction
versus PPARγ activation, provides a molecular basis for using
troglitazone as scaffold to develop novel agents with high
potency in repressing cyclin D1 expression. This structural
optimization has led to a novel class of cyclin D1-ablative agents
with low-micromolar potency.
Overexpression of the cell cycle control gene cyclin D1
represents a key mechanism underlying tumorigenesis, tumor
progression, and metastasis in a variety of human cancers.1-6
Evidence indicates that cyclin D1 regulates cell cycle progres-
sion and proliferation through the interaction with multiple
signaling targets.5,6 Cyclin D1 serves as the regulatory subunit
of cyclin-dependent kinases (CDKs) 4 and 6 and exhibits the
ability to bind and sequester the CDK inhibitor p27.5,6 Together,
these functions facilitate CDK-mediated inactivating phospho-
rylation of the retinoblastoma protein (pRb), resulting in G1/S
progression. In addition, this cell cycle regulator may regulate
transcription through physical associations with a plethora of
transcriptional factors, coactivators, and corepressors that govern
histone acetylation and chromatin remodeling proteins.6 The
concerted action of these CDK-dependent and CDK-independent
functions underscores the oncogenic potential of cyclin D1 in
many forms of cancer.7 Inhibition of cyclin D1 expression by
antisense oligonucleotides has been shown to block tumorigen-
esis or to reverse the transformed phenotype of human esoph-
ageal,8 lung,9 colon,10 pancreatic,11 gastric,12,13 melanoma,14 and
squamous cancer cells15 in mice.
In light of the function of cyclin D1 as an oncogene, targeting
its expression represents a promising strategy for cancer
therapy.16 Cyclin D1 is induced by a multitude of proliferative
and transforming signaling pathways, including the RAS/RAF/
ERK, PI3K/PDK-1, WNT/â-catenin, and PKA/CREB pathways,
and undergoes rapid proteasomal degradation upon the with-
drawal of growth factors.5,6 Previously, we reported that the
peroxisome proliferator-activated receptor γ (PPARγ) agonist
troglitazone, at high doses, facilitated the proteasomal proteolysis
of cyclin D1 independently of PPARγ.17 The dissociation of
these two pharmacological activities, i.e., cyclin D1 repression
Chemistry
In our previous report, we demonstrated that introduction of
a double bond adjoining the terminal thiazolidine-2,4-dione ring
of troglitazone abrogated the PPARγ agonist activity without
compromising the cyclin D1-repressing activity.17 Thus, the
resulting PPAR-γ-inactive analogue ∆2TG (Figure 1A) was
used as a lead compound in this study for structural modifica-
tions. To envisage the role of the vitamin E antioxidant moiety
in the effect of ∆2TG on cyclin D1 ablation, different substit-
uents were added to the terminal hydroxyl function of ∆2TG,
yielding compounds 1-5. Among them, addition of an allyl,
3,3-dimethylallyl, or benzyl group to ∆2TG resulted in a
multifold increase in the cyclin D1-ablating activity. Subse-
quently, compound 6 was synthesized to examine the effect of
the orientation of the thiazolidinedione ring on the activity.
Furthermore, we used the allyl- and 3,3-dimethylallyl-substituted
analogues of ∆2TG as platforms to carry out structural
optimization, which generated two series of compounds, i.e.,
7-10 and 11-16, respectively. These ∆2TG derivatives were
synthesized according to a general procedure described in Figure
1B. All derivatives synthesized, including ∆2TG, were racemic
compounds unless otherwise indicated.
Results
Pharmacological Exploitation of PPARγ-Independent Ef-
fect of Troglitazone on Cyclin D1 Repression To Generate
a Novel Class of Cyclin D1-Ablative Agents. Dose-dependent
effect of troglitazone and ∆2TG on cyclin D1 repression in
MCF-7 cells was examined by Western blot analysis after 72 h
of drug treatment (Figure 2A). ∆2TG, though devoid of PPARγ
agonist activity, exhibited slightly higher potency in repressing
* To whom correspondence should be addressed. Phone: (614) 688-
4008. Fax: (614) 688-8556. E-mail: chen.844@osu.edu.
† The Ohio State University.
‡ National Taiwan University and National Health Research Institutes.
10.1021/jm060057h CCC: $33.50 © 2006 American Chemical Society
Published on Web 06/29/2006

NoVel Cyclin D1-AblatiVe Agents
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 15 4685
Figure 1. (A) Representative structures of troglitazone and derivatives
and (B) the general synthetic procedure for troglitazone derivatives.
cyclin D1 compared to troglitazone. For example, the concentra-
tion required for the complete abrogation of cyclin D1 expres-
sion was approximately 30 µM for ∆2TG vis-a`-vis 40 µM for
troglitazone. Earlier, we have demonstrated that this troglitazone-
and ∆2TG-mediated cyclin D1 down-regulation was caused by
ubiquitin-dependent proteasomal degradation.17 Together, these
findings suggest that the effect of these agents on cyclin D1
repression was independent of PPARγ activation. Consequently,
∆2TG was used as a starting point for structural optimization
to generate cyclin D1-ablative agents with low-micromolar
potency.
Lead Optimization: Steric and Stereochemical Effects.
∆2TG was first conjugated with different substituents through
an ether linkage at the terminal hydroxyl function, generating
compounds 1-5. These derivatives were analyzed by Western
blotting for their ability, within the dose range of 2.5-7.5 µM,
to repress cyclin D1 expression in MCF-7 cells (Figure 2B).
Among the five derivatives examined, the allyl (1), 3,3-
dimethylallyl (2), and benzyl (3) derivatives gave robust
reduction in cyclin D1 levels at concentrations below 5 µM after
72 h of exposure, while the cinnamyl (4) and succinyl (5)
counterparts had only moderate effect even at 7.5 µM. This
differential increase in potency suggests a unique mode of ligand
recognition that preferred less bulky hydrophobic side chains.
The 4- to 5-fold improvement in the cyclin D1-ablating activity
of compounds 1-3 over ∆2TG indicates the importance of this
hydrophobic interaction in ligand binding.
Figure 2. Dose-dependent effect of troglitazone, ∆2TG, and com-
pounds 1-16 on cyclin D1 ablation in MCF-7 breast cancer cells.
MCF-7 cells were exposed to the individual agents at the indicated
concentrations in 5% FBS-supplemented medium for 72 h, and the
expression of cyclin D1 was analyzed by Western blot analysis with
â-actin as an internal standard: (A) Western blot analysis of the dose-
dependent effect of troglitazone and ∆2TG on cyclin D1 repression;
(B) a summary of the differential effect of compounds 1-16 on cyclin
D1 ablation. The general structures for these compounds are shown in
Figure 1. Intensities of cyclin D1 immunostaining in Western blots
were measured by densitometry and after being normalized to that of
actin were presented as numerical data with the DMSO control as 100%.
(C) Representative Western blotting of compounds 2, 6, 8, and 13 is
shown at the indicated concentrations.
exhibited the highest potency with multifold improvement over
the parent compound, followed by the methyl derivative (9),
while the bromo and methyoxy counterparts (7 and 10,
respectively) showed modest or no appreciable reduction in
cyclin D1-ablating activity (Figure 2B). The IC50 of 8 in
mediating the repression of cyclin D1 was approximately 2.5
µM after 72 h of exposure, compared to greater than 5 µM for
1. On the other hand, derivatization of 2 with various functional
groups (compounds 11-16) led to a decrease in potency
irrespective of the size and stereoelectronic property of these
substituents (Figure 2B).
In light of the high potency of 8, we further compared the
activity of its optically active isomers. Figure 3 depicts the dose-
dependent effect of (S)-8 (left panel, named STG28 from this
point on) and (R)-8 (right panel) in repressing the expression
of cyclin D1 in MCF-7 cells after a 24-h treatment. These data
clearly indicate a stereochemical preference in the pharmaco-
logical action, i.e., the potency of the (S)-isomer was severalfold
higher than that of the (R)-counterpart. In line with its enhanced
ability in cyclin D1 ablation, STG28 exhibited significantly
Our data also suggest the requirement of a linear, elongated
ligand structure for the high potency in cyclin D1 ablation. For
example, changing the relative position of the terminal meth-
ylene-2,4-thiazolidinedione moiety to the phenyl ring from para
to meta (1 f 6) substantially attenuated the cyclin D1-repressing
potency (Figure 2B).
With 1 and 2 as lead compounds, we further examined the
effect of modifying the benzylidene moiety with different
substitutions on cyclin D1-ablating activity. Among the four
different analogues of 1 examined, the methoxy derivative (8)

4686 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 15
Huang et al.
Figure 3. Dose-dependent effect of (S)-8 (STG28) versus (R)-8 on
cyclin D1 ablation. MCF-7 cells were exposed to the individual agents
at the indicated concentrations in 5% FBS-supplemented medium for
24 h, and the expression of cyclin D1 was analyzed by Western blot
analysis with â-actin as an internal standard.
Figure 5. Evidence that the effect of unsaturated thiazolidinedione
derivatives on cyclin D1 repression is reversible. The effect of the
unsaturated thiazolidinedione 17 vis-a`-vis its saturated counterpart 18,
each at 10 µM, on cyclin D1 repression in MCF-7 cells was examined
at different intervals throughout a 48 h period in two different manners.
(A) For continuous exposure, cells in T-75 flasks were incubated in
drug-containing, 5% FBS-supplemented medium for 48 h. (B) For
washout at 24 h of treatment, cells in T-75 flasks were exposed to the
agent for 24 h, followed by incubation in drug-free medium for an
additional 24 h. Cyclin D1 levels in cell lysates were analyzed by
Western blot analysis with â-actin as an internal standard.
cell cycle-regulatory proteins, underscoring the high degree of
specificity in the effect of STG28 on cyclin D1 repression.
Second, semiquantitative PCR (polymerase chain reaction)
showed that the mRNA level of cyclin D1 remained unaltered
after 24 h of exposure to different doses of STG28 (Figure 4B),
indicating that the repression occurred at the protein level.
Third, we tested the effect of the proteasome inhibitor MG132
on the protection of STG28-mediated cyclin D1 ablation (Figure
4C). As shown, MG132 at 5 and 10 µM was effective in
rescuing the drug-induced cyclin D1 repression, providing a
mechanistic link with proteasome-mediated proteolysis.
Together, these findings confirmed that STG28 mediated the
repression of cyclin D1 through the same mode of mechanism
as that of troglitazone and ∆2TG. The target that these ∆2TG
derivatives use to facilitate proteasomal degradation of cyclin
D1 remains unclear. However, the presence of the benzylidine-
thiazolidine-2,4-dione substructure raised a possibility that these
∆2TG derivatives might act as “Michael acceptors” by co-
valently modifying the target enzyme/protein upon binding. To
examine this possibility, we assessed the effect of an unsaturated
thiazolidinedione (17) vis-a`-vis its saturated counterpart (18)
on cyclin D1 ablation by using a washout experiment. Com-
pounds 17 and 18 exhibited IC50 of 6 and 9 µM, of which the
relative ratio was in line with that of ∆2TG to troglitazone.
Figure 5A shows the time-dependent effect of 17 (left panel)
and 18 (right panel), each at 10 µM, on the suppression of cyclin
D1 expression. In a parallel experiment in which these agents
were washed out from the medium after 24 h of exposure, the
cyclin D1 level rapidly returned to normal regardless of the
unsaturated thiazolidinedione moiety (Figure 5B). These data
suggest a reversible nature of this ligand-protein interaction.
Figure 4. Mechanistic validation of the mode of action of STG28.
(A) Effect of STG28 on cyclin D1 ablation is highly specific. MCF-7
cells were exposed to STG28 at the indicated concentrations in 5%
FBS-supplemented medium for 24 h, and the expression of various
cyclins, CDKs, the CDK inhibitors p21 and p27 was analyzed by
Western blot analysis with â-actin as an internal standard. (B) STG28-
facilitated cyclin D1 ablation is mediated at the post-transcriptional
level. Drug-treated MCF-7 cells were subjected to total RNA isolation,
followed by RT-PCR analysis of mRNA transcripts of cyclin D1 gene,
as described in the Experimental Section. (C) STG28-facilitated cyclin
D1 ablation is mediated through proteasomal degradation. MCF-7 cells
were exposed to 10 µM STG28 in the presence of various concentrations
of the proteasome inhibitor MG132 in 5% FBS-supplemented medium
for 24 h, and the expression of cyclin D1 was analyzed by Western
blot analysis with â-actin as an internal standard.
higher potency than ∆2TG in inhibiting MCF-7 cell proliferation
(IC50 of 5 vs 55 µM).
Mechanistic Validation of the Mode of Action of STG28.
Previously, we demonstrated that troglitazone and ∆2TG
facilitated proteasomal degradation of cyclin D1 in a highly
specific manner.17 Here, we carried out the following three
experiments to validate the mechanism by which STG28
mediated cyclin D1 ablation. First, to demonstrate that the
ablative effect of STG28 was unique to cyclin D1, we assessed
the expression levels of cyclins D2, D3, A, B, and E, CDKs 2
and 4, and the CDK inhibitors p21 and p27 in MCF-7 cells
exposed to different doses of STG28 for 24 h (Figure 4A). As
shown, with the exception of a slight decrease in p21 levels,
no appreciable change was noted in the expression level of these
Discussion
On the basis of the mechanistic finding that the effect of
troglitazone on cyclin D1 repression was independent of PPARγ
activation,17 this study embarked on the structural optimization
of troglitazone to develop a novel class of cyclin D1-ablative
agents. Our premise that troglitazone’s pharmacological activi-
ties in cyclin D1 ablation and PPARγ activation could be
structurally dissociated was borne out by ∆2TG, which, albeit

NoVel Cyclin D1-AblatiVe Agents
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 15 4687
(Z)-5-{4-[(6-Allyloxy-2,5,7,8-tetramethylchroman-2-yl)methoxy]-
benzylidene}thiazolidine-2,4-dione (1). Step a. (1) To a stirring
solution of 0.78 g of LiAlH4 (20 mmol) in 10 mL of THF at 0 °C
was added 5 g (20 mmol) of 6-hydroxy-2,5,7,8-tetramethylchroman-
2-carboxylic acid (Sigma-Aldrich, St. Louis, MO) in 250 mL of
THF dropwise over a period of 1 h. The solution was stirred at
room temperature under nitrogen for another 6 h, and 4 mL of 0.25
N NaOH was added to quench the reaction. The mixture was stirred
at room temperature for 1 h, filtered, and concentrated under
reduced pressure. The resulting residue was purified by flash silica
gel chromatography (ethyl acetate/hexane, 1:2) to afford 2-hy-
droxymethyl-2,5,7,8-tetramethylchroman-6-ol in 75% yield.
(2) A mixture of the above purified product (2 mmol), allyl
bromide (2 mmol), and K2CO3 (3 mmol) in 10 mL of acetone was
refluxed for 48 h, filtered, and concentrated. The resulting residue
was purified by flash silica gel chromatography (ethyl acetate/
hexanes, 1:5) to yield (6-allyloxy-2,5,7,8-tetramethylchroman-2-
yl)methanol in 81% yield.
(3) Triflate anhydride (0.19 mL, 1.1 mmol) was added dropwise
into the mixture of the above purified product (1 mmol), pyridine
(1.2 mmol), and 5 mL of CH2Cl2 at 0 °C. The solution was stirred
at 0 °C for 30 min and concentrated. The resulting residue was
purified by flash silica gel chromatography (ethyl acetate/hexanes,
1:10), giving (6-allyloxy-2,5,7,8-tetramethylchroman-2-yl)methyl
trifluoromethanesulfonate in 92% yield.
Step b. The purified triflate from step a (0.5 mmol) in 2 mL of
DMF was treated with K2CO3 (0.5 mmol) in 10 mL of DMF at
0 °C for 30 min, followed by 4-hydroxybenzaldehyde (0.5 mL,
4.7 mmol). The mixture was stirred at room temperature overnight,
poured into water to quench the reaction, extracted with ethyl
acetate, and concentrated. The resulting residue was purified by
flash silica gel chromatography (ethyl acetate/hexanes, 1:10),
resulting in 4-[(6-allyloxy-2,5,7,8-tetramethylchroman-2-yl)methoxy]-
benzaldehyde in 93% yield.
devoid of PPARγ activity, exhibited a slightly higher cyclin
D1-ablating activity. Presumably, the rigid, liner substructure
of benzylidene-thiazolidinedione of ∆2TG prohibited its interac-
tion with PPARγ. Subsequent modifications of ∆2TG by adding
hydrophobic side chains, including allyl, dimethylallyl, and
benzyl, to the terminal hydroxyl function (compounds 1, 2, and
3, respectively) resulted in a multifold increase in cyclin D1-
repressing potency. This increase, however, was not noted with
an extended aromatic side chain such as cinnamyl (4) or a
charged group such as succinyl (5). Moreover, the requirement
of a linear structure for the ligand recognition was underscored
by the substantially lower potency of 6 compared to 1. Further
modifications of the benzylidene moiety in 1 and 2 revealed a
subtle effect of the 3-substitution on cyclin D1 ablation. Among
the 10 derivatives examined (7-16), 8 showed the highest
activity, followed by 9, while the other derivatives displayed
decreased potency. It is also noteworthy that the ability of 8 in
cyclin D1 repression was stereospecific; i.e., the (S)-enantiomer
(STG28) was more active than the (R)-counterpart. STG28 was
able to repress cyclin D1 at low-micromolar concentrations,
while (R)-8 required at least 10 µM to achieve the same extent
of repression.
Equally important, STG28-mediated cyclin D1 ablation was
highly specific. Our data indicate that none of the other cyclins
examined were affected by STG28. The mode of action of
STG28 in facilitating cyclin D1 ablation was identical to that
of its parent compound, i.e., via proteasome-mediated proteoly-
sis. From a mechanistic perspective, STG28 provides a useful
pharmacological tool to study the ubiquitin-dependent protea-
somal degradation of cyclin D1, which constitutes the focus of
our current investigation.
Step c. A mixture of the above purified aldehyde (0.3 mmol),
4-thiazolidinedione (0.6 mmol), and a catalytic amount of piperidine
was refluxed in 5 mL of EtOH for 24 h and concentrated. The oil
product was dissolved in ethyl acetate, poured into water, and
acidified with acetic acid. The mixture was extracted with ethyl
acetate and concentrated. The resulting residue was purified by flash
silica gel chromatography (ethyl acetate/hexanes, 1:5), generating
1 in 71% yield: 1H NMR (CDCl3) δ 1.30 (s, 3H), 1.83-1.93 (m,
1H), 2.04-2.16 (m, 10H), 2.58 (t, J ) 5.8 Hz, 2 H), 3.95 (d, J )
8.8 Hz, 1 H), 4.03 (d, J ) 8.8 Hz, 1 H), 4.15 (d, J ) 11.5 Hz, 2
H), 5.23 (d, J ) 10.1 Hz, 1 H), 5.41 (d, J ) 18.2 Hz, 1 H), 6.02-
6.17 (m, 1 H), 7.10 (d, J ) 9.8 Hz, 2 H), 7.48 (d, J ) 9.8 Hz, 2
H), 7.79 (s, 1 H), 8.56 (s, 1 H); HRMS exact mass of (M + Na)+,
502.1658 amu; observed mass of (M + Na)+, 502.1685 amu. Anal.
(C27H29NO5S) C, H, N.
Conclusion
STG28 represents the first small-molecule agent that exhibits
high potency in repressing cyclin D1 expression. This agent
may have therapeutic relevance in light of the pivotal role of
cyclin D1 in promoting tumorigenesis and drug resistance. In
breast cancer, overexpression of this cell cycle regulator confers
resistance to antiestrogens and represents a negative predictive
factor for tamoxifen response. Thus, these findings underscore
the potential use of this anti-cyclin D1 therapy in the treatment
of breast cancer. Testing of STG28 in an in vivo breast tumor
model is currently under way.
Experimental Section
(Z)-5-(4-((2,5,7,8-Tetramethyl-6-(3-methylbut-2-enyloxy)chro-
man-2-yl)methoxy)benzylidene)thiazolidine-2,4-dione (2): 1H
NMR (CDCl3) δ 1.40 (s, 3 H), 1.63 (s, 3 H), 1.69 (s, 3 H), 1.88-
1.93 (m, 1 H), 2.04-2.18 (m, 10 H), 2.61 (t, J ) 6.8, 2 H), 3.95
(d, J ) 9.3 Hz, 1 H), 4.03 (d, J ) 9.3 Hz, 1 H), 4.15 (d, J ) 7.0,
2 H), 5.58 (t, J ) 7.2, 1 H), 7.01 (d, J ) 8.8 Hz, 2 H), 7.42 (d, J
) 8.8 Hz, 2 H), 7.90 (s, 1 H), 8.84 (s, 1 H); HRMS exact mass of
(M + Na)+, 530.1971 amu; observed mass of (M + Na)+, 530.1975
amu. Anal. (C29H33NO5S) C, H, N.
Chemical reagents and organic solvents were purchased from
Aldrich unless otherwise mentioned. Nuclear magnetic resonance
spectra (1H NMR) were measured on a Bruker AC 300, Bruker
DPX 300, or Bruker DRX 400 model spectrometer. Chemical shifts
(δ) were reported in parts per million (ppm) relative to the TMS
peak. Electrospray ionization (ESI) mass spectrometry analyses were
performed with a Micromass QTOF electrospray mass spectrometer.
Elemental analyses were performed by the Atlantic Microlab, Inc.
(Norcross, GA).
(Z)-5-{4-[(6-Benzyloxy-2,5,7,8-tetramethylchroman-2-yl)-
methoxy]-3-bromobenzylidene}thiazolidine-2,4-dione (3): 1H
NMR (CDCl3) δ 1.42 (s, 3 H), 1.84-1.95 (m, 1 H), 2.04-2.20
(m, 10 H), 2.62 (t, J ) 6.9 Hz, 2 H), 3.95 (d, J ) 9.3, 1 H), 4.01
(d, J ) 9.3 Hz, 1 H), 4.67 (s, 2 H), 7.01 (d, J ) 8.9 Hz, 2 H),
7.35-7.50 (m, 7 H), 7.79 (s, 1 H), 8.69 (brs, 1 H); HRMS exact
mass of (M + Na)+, 552.1815 amu; observed mass of (M + Na)+,
552.1839 amu. Anal. (C31H31NO5S) C, H, N.
(Z)-5-{4-[(6-Cinnamyloxy-2,5,7,8-tetramethylchroman-2-yl)-
methoxy]benzylidene}thiazolidine-2,4-dione (4): 1H NMR (CDCl3)
δ 1.41 (s, 3 H), 1.83-1.94 (m, 1 H), 2.02-2.20 (m, 10 H), 2.62 (t,
J ) 6.0 Hz, 2 H), 3.95 (d, J ) 8.9 Hz, 1 H), 4.04 (d, J ) 8.9 Hz,
1 H), 4.34 (d, J ) 5.8 Hz, 2 H), 6.40-6.51 (m, 1 H), 6.73 (d, J )
Troglitazone and the proteasome inhibitor MG132 were pur-
chased from Sigma-Aldrich (St. Louis, MO). ∆2TG was synthesized
as previously described.17 Mouse anti-cyclin D1 was purchased from
Cell Signaling Technology Inc. (Beverly, MA). Rabbit antibodies
against CDK2, CDK4, cyclin A, cyclin B, cyclin D2, cyclin D3,
cyclin E, p21, p27 were purchased from Santa Cruz Biotechnology,
Inc. (Santa Cruz, CA). Mouse monoclonal anti-actin was purchased
from MP Biomedicals (Irvine, CA). MTT [3-(4,5-dimethylthiazol-
2-yl)-2,5-diphenyl-2H-tetrazolium bromide] for cell viability assay
was purchased form TCI America, Inc. (Portland, OR).
Compounds 1-18 were synthesized according to the general
procedure described in Figure 1B, which is illustrated by the
synthesis of 1, described as follows.

4688 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 15
Huang et al.
16 Hz, 1 H), 7.01 (d, J ) 8.7 Hz, 2 H), 7.25-7.50 (d, 7 H), 7.79
(s, 1 H), 8.54 (brs, 1 H); HRMS exact mass of (M + Na)+, 578.1971
amu; observed mass of (M + Na)+, 578.1966 amu.
amu; observed mass of (M + Na)+, 608.1094 amu. Anal. (C29H32-
BrNO5S) C, H, N.
(Z)-5-(3-Methoxy-4-{[2,5,7,8-tetramethyl-6-(3-methylbut-2-
enyloxy)chroman-2-yl]methoxy}benzylidene)thiazolidine-2,4-di-
one (13): 1H NMR (CDCl3) δ 1.42 (s, 3 H), 1.69 (s, 3 H), 1.78 (s,
3 H), 1.86-1.93 (m, 1 H), 2.01-2.17 (m, 10 H), 2.61 (t, J ) 6.8
Hz, 2 H), 3.87 (s, 3 H), 4.00 (d, J ) 9.5 Hz, 1 H), 4.09 (d, J ) 9.5
Hz, 1 H), 4.14 (d, J ) 6.8 Hz, 2 H), 5.58 (t, J ) 7.1 Hz, 1 H),
6.96-7.06 (m, 3 H), 7.76 (s, 1 H); HRMS exact mass of (M +
Na)+, 560.2077; observed mass of (M + Na)+, 560.2086.
(Z)-5-(3-Methyl-4-{[2,5,7,8-tetramethyl-6-(3-methylbut-2-eny-
loxy)chroman-2-yl]methoxy}benzylidene)thiazolidine-2,4-di-
one (14): 1H NMR (CDCl3) δ 1.44 (s, 3 H), 1.69 (s, 3 H), 1.78 (s,
3 H), 1.86-1.94 (m, 1 H), 2.02-2.20 (m, 10 H), 2.26 (s, 3 H),
2.61 (t, J ) 3.2 Hz, 2 H), 3.95 (d, J ) 9.3 Hz, 1 H), 4.03 (d, J )
9.3 Hz, 1 H), 4.17 (d, J ) 6.9 Hz, 2 H), 5.58 (t, J ) 5.5 Hz, 1 H),
6.88 (d, J ) 9.2 Hz, 1H), 7.27 (s, 1 H), 7.29 (d, J ) 9.2 Hz, 1H),
7.76 (s, 1 H), 8.24 (brs,1 H); HRMS exact mass of (M + Na),
544.2128 amu; observed mass of (M + Na)+, 544.2134 amu.
(Z)-5-(3-Ethoxy-4-{[2,5,7,8-tetramethyl-6-(3-methylbut-2-eny-
loxy)chroman-2-yl]methoxy)benzylidene}thiazolidine-2,4-di-
(Z)-4-[2-({4-[(2,4-Dioxothiazolidin-5-ylidene)methyl]phenoxy}-
methyl)-2,5,7,8-tetramethylchroman-6-yloxy]-4-oxobutanoic Acid
(5): 1H NMR (CDCl3) δ 1.36 (s, 3 H), 1.81-2.10 (m, 11 H), 2.62
(m, 2 H), 2.77-2.82 (m, 2 H), 2.88-2.92 (m, 2 H), 3.95 (d, J )
9.2, 1 H,), 4.02 (d, J ) 9.2 Hz, 1 H), 6.96 (d, J ) 6.6 Hz, 2 H),
7.40 (d, J ) 6.6 Hz, 2 H), 7.77 (s, 1 H), 8.43 (brs, 1 H); HRMS
exact mass of (M + Na)+, 562.1506 amu; observed mass of (M +
Na)+, 562.1480.
(Z)-5-{3-[(6-Allyloxy-2,5,7,8-tetramethylchroman-2-yl)methoxy]-
benzylidene}thiazolidine-2,4-dione (6): 1H NMR (CDCl3) δ 1.35
(s, 3 H), 1.75-1.98 (m, 1 H), 2.04-2.22 (m, 10 H), 2.61 (t, J )
5.9 Hz, 2 H), 3.95 (m, 2 H), 4.15 (d, J ) 11.5 Hz, 2 H), 5.25 (d,
J ) 8.6 Hz, 1 H), 5.43 (d, J ) 14.3 Hz, 1 H), 6.02-6.17 (m, 1 H),
7.01 (m, 2 H), 7.08 (m, 1 H), 7.35 (t, J ) 6.9 Hz, 1 H), 7.75 (s, 1
H), 8.38 (s, 1 H); HRMS exact mass of (M + Na)+, 502.1659 amu;
observed mass of (M + Na)+, 502.1677.
(Z)-5-{4-[(6-Allyloxy-2,5,7,8-tetramethylchroman-2-yl)methoxy]-
3-bromobenzylidene}thiazolidine-2,4-dione (7): 1H NMR (CDCl3)
δ 1.47 (s, 3 H), 1.94-2.19 (m, 11 H), 2.60-2.64 (m, 2 H), 4.00
(d, J ) 9.2 Hz, 1 H), 4.07 (d, J ) 9.2 Hz, 1 H), 4.16 (d, J ) 5.5
Hz, 2 H), 5.23 (d, J ) 10.4 Hz, 1 H), 5.41 (d, J ) 17.2 Hz, 1 H),
5.98-6.11 (m, 1 H), 6.96 (d, J ) 8.7 Hz, 1 H), 7.36 (dd, J ) 8.7,
2.1 Hz, 1 H), 7.67 (d, J ) 2.1 Hz, 1 H), 7.70 (s, 1 H), 8.30 (brs,
1 H); HRMS exact mass of (M + Na)+, 580.0763 amu; observed
mass of (M + Na)+, 580.0763 amu.
(Z)-5-{4-[(6-Allyloxy-2,5,7,8-tetramethylchroman-2-yl)methoxy]-
3-methoxybenzylidene}thiazolidine-2,4-dione (8): 1H NMR (CDCl3)
δ 1.42 (s, 3 H), 1.88-1.92 (m, 1 H), 2.04-2.18 (m, 10 H), 2.61 (t,
J ) 7.2 Hz, 2 H), 3.88 (s, 3 H), 3.98 (d, J ) 9.9 Hz, 1 H), 4.09 (d,
J ) 9.9 Hz, 1 H), 4.16 (d, J ) 5.5 Hz, 2 H), 5.23 (d, J ) 10.4 Hz,
1 H), 5.40 (d, J ) 17.2 Hz, 1 H), 6.03-6.15 (m, 1 H), 6.96-7.06
(m, 3 H), 7.76 (s, 1 H), 8.34 (s, 1 H); HRMS exact mass of (M +
Na)+, 532.1764 amu; observed mass of (M + Na)+, 532.1762 amu.
Anal. (C28H31NO6S) C, H, N.
(Z)-5-{4-[(6-Allyloxy-2,5,7,8-tetramethylchroman-2-yl)methoxy]-
3-methylbenzylidene}thiazolidine-2,4-dione (9): 1H NMR (CDCl3)
δ 1.43 (s, 3 H), 1.89-1.94 (m, 1 H), 2.05-2.20 (m, 10 H), 2.26
(s, 3 H), 2.61 (t, J ) 6.5 Hz, 2 H), 3.95 (d, J ) 9.3 Hz, 1 H), 4.02
(d, J ) 9.3 Hz, 1 H), 4.16 (d, J ) 5.5 Hz, 2 H), 5.23 (d, J ) 10.4
Hz, 1 H), 5.41 (d, J ) 17.3 Hz, 1 H), 6.04-6.15 (m, 1 H), 6.88 (d,
J ) 8.8 Hz, 1 H), 7.27 (s, 1 H), 7.29 (d, J ) 8.0 Hz, 1 H), 7.76 (s,
1 H); HRMS exact mass of (M + Na)+, 516.1815 amu; observed
mass of (M + Na)+, 516.1837 amu. Anal. (C28H31NO5S) C, H, N.
(Z)-5-{4-[(6-Allyloxy-2,5,7,8-tetramethylchroman-2-yl)methoxy]-
3-ethoxybenzylidene}thiazolidine-2,4-dione (10): 1H NMR (CDCl3)
δ 1.45 (s, 3 H), 1.58 (s, 3 H), 1.85-1.95 (m, 1 H), 2.02 (s, 3 H),
2.12-2.25 (m, 7 H), 2.60 (m, 2 H), 3.97-4.31 (m, 6 H), 5.23 (d,
J ) 8.6 Hz, 1 H), 5.40 (d, J ) 13.3 Hz, 1 H), 6.05-6.16 (m, 1 H),
6.94-7.10 (m, 3 H), 7.76 (s, 1 H), 8.31 (s, 1 H); HRMS exact
mass of (M + Na)+, 546.1921 amu; observed mass of (M + Na)+,
546.1907 amu.
(Z)-5-(3-Fluoro-4-{[2,5,7,8-tetramethyl-6-(3-methylbut-2-eny-
loxy)chroman-2-yl]methoxy}benzylidene)thiazolidine-2,4-di-
one (11): 1H NMR (CDCl3) δ 1.42 (s, 3 H), 1.69 (s, 3 H), 1.78 (s,
3 H), 1.87-1.93 (m, 1 H), 1.96 (s, 3 H), 2.01-2.17 (m, 7 H), 2.61
(t, J ) 6.5 Hz, 2 H), 4.02 (d, J ) 9.5 Hz, 1 H), 4.10 (d, J ) 9.5
Hz, 1 H), 4.15 (d, J ) 7.2 Hz, 2 H), 5.57 (t, J ) 6.9 Hz, 1 H), 7.05
(t, J ) 8.3 Hz, 1 H), 7.17-7.23 (m, 2 H), 7.72 (s,1 H), 8.46 (brs,
1 H); HRMS exact mass of (M + Na)+, 548.1877 amu; observed
mass of (M + Na)+, 548.1889 amu.
1
one (15): H NMR (CDCl3) δ 1.40-1.46 (m, 6 H), 1.70 (s, 3 H),
1.79 (s, 3 H), 1.87-1.93 (m, 1 H), 2.01 (s, 3 H), 2.09-2.17 (m, 7
H), 2.62 (t, J ) 7.0 Hz, 2 H), 3.97-4.16 (m, 6 H), 5.58 (t, J ) 7.1
Hz, 1 H), 6.97-7.08 (m, 3 H), 7.76 (s, 1 H), 8.59 (s, 1 H); HRMS
exact mass of (M + Na), 574.2342; observed mass of (M + Na)+,
574.2122.
(Z)-5-[(4′-Fluoro-6-{[2,5,7,8-tetramethyl-6-(3-methylbut-2-
enyloxy)chroman-2-yl]methoxy}biphenyl-3-yl)methylene]thiazo-
1
lidine-2,4-dione (16): H NMR (CDCl3) δ 1.27 (s, 3 H), 1.69-
1.94 (m, 8 H), 2.00 (s, 3 H), 2.10 (s, 3 H), 2.17 (s, 3 H), 2.46-
2.61 (m, 2 H), 3.99 (s, 2 H), 4.14 (d, J ) 7.0 Hz, 2 H), 5.58 (t, J
) 7.1 Hz, 1 H), 7.02-7.10 (m, 3 H), 7.41-7.52 (m, 4 H), 7.81 (s,
1 H), 8.30 (brs, 1 H); HRMS exact mass of (M + Na)+, 624.2190;
observed mass of (M + Na)+, 624.2206.
(Z)-5-{4-[2-(6-Butoxy-2,5,7,8-tetramethyl-chroman-2-yl)ethyl]-
3-trifluoromethylbenzylidene}thiazolidine-2,4-dione (17): 1H
NMR (CDCl3) δ 0.93 (t, J ) 7.0 Hz, 3 H), 1.31-1.54 (m, 5 H),
1.73-1.84 (m, 2 H), 1.87-1.95 (m, 1 H), 1.95-2.19 (m, 10 H),
2.60 (m, 2 H), 3.62 (t, J ) 6.7 Hz, 2 H), 4.03 (q, J ) 9.0 Hz, 1 H),
4.08 (q, J ) 9.3 Hz, 1 H), 7.09 (d, J ) 9.0 Hz, 1 H), 7.59 (dd, J
) 9.0, 2.1 Hz, 1 H), 7.70 (d, J ) 2.1 Hz, 1H), 7.79 (s, 1 H), 8.83
(s, 1 H); HRMS exact mass of (M + Na)+, 584.2058; observed
mass of (M + Na)+, 584.2094.
5-{4-[2-(6-Butoxy-2,5,7,8-tetramethylchroman-2-yl)ethyl]-3-
trifluoromethylbenzyl}thaizolodine-2,4-dione (18): 1H NMR
(CDCl3) δ 0.99 (t, J ) 7.2, 3 H), 1,34-1.68 (m, 5 H), 1.69-1.86
(m, 2 H), 1.86-1.98 (m, 1 H), 1.98-2.22 (m, 10 H), 2.62 (1 H),
3.18 (m, 1 H), 3.46 (m, 1 H), 3.64 (t, J ) 6.9 Hz, 2 H), 4.26 (q, J
) 10 Hz, 2 H), 4.51 (m, 1 H), 6.95 (d, J ) 6.6 Hz, 1H), 7.26 (d,
J ) 6.6 Hz, 1 H), 7.44 (s, 1 H), 8.165 (s, 1 H); HRMS exact mass
of (M + Na)+, 586.2215; observed mass of (M + Na)+, 586.2237.
Cell Culture. ER-positive MCF-7 breast cancer cells were
obtained from the American Type Culture Collection (Manassas,
VA) and were maintained in DMEM/Ham’s F-12 medium supple-
mented with 10% fetal bovine serum (FBS, Gibco) at 37 °C in a
humidified incubator containing 5% CO2.
Western Blot Analysis. MCF-7 cells were seeded in 10% FBS-
containing DMEM/Ham’s F-12 medium for 24 h and treated with
various agents as indicated. After individual treatments for 24 h,
both the incubation medium and adherent cells in T-25 or T-75
flasks were scraped and collected by centrifugation at 2200 rpm
for 10 min. The supernatants were recovered, placed on ice, and
triturated with 20-50 µL of a chilled lysis buffer (M-PER
mammalian protein extraction reagent; Pierce, Rockford, IL), to
which was added 1% protease inhibitor cocktail (set III, EMD
Biosciences, Inc.; San Diego, CA). After a 30-min incubation on
ice, the mixture was centrifuged at 16100g for 3 min. An amount
of 2 µL of the suspension was taken for protein analysis using the
Bradford assay kit (Bio-Rad, Hercules, CA). To the remaining
solution was added the same volume of 2× SDS-polyacrylamide
(Z)-5-(3-Bromo-4-{[2,5,7,8-tetramethyl-6-(3-methylbut-2-eny-
loxy)chroman-2-yl)methoxy]benzylidene}thiazolidine-2,4-di-
one (12): 1H NMR (CDCl3) δ 1.40 (s, 3 H), 1.69 (s, 3 H), 1.78 (s,
3 H), 1.91-1.97 (m, 1 H), 2.02-2.21 (m, 10 H), 2.61 (m, 2 H),
3.94 (d, J ) 9.5 Hz, 1H), 4.03 (d, J ) 9.5 Hz, 1 H), 4.14 (d, J )
6.8 Hz, 2 H), 5.58 (t, J ) 7.1 Hz, 1 H), 6.96 (d, J ) 8.6 Hz, 1 H),
7.38 (dd, J ) 8.6, 2.0 Hz, 1 H), 7.67 (d, J ) 2.0 Hz, 1 H), 7.79
(s,1 H), 8.56 (brs, 1 H); HRMS exact mass of (M + Na)+, 608.1076

NoVel Cyclin D1-AblatiVe Agents
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 15 4689
References
gel electrophoresis sample loading buffer (100 mM Tris-HCl, pH
6.8, 4% SDS, 5% â-mercaptoethanol, 20% glycerol, and 0.1%
bromphenol blue). The mixture was boiled for 10 min. Equal
amounts of proteins were loaded onto 10% SDS-polyacrylamide
gels. After electrophoresis, protein bands were transferred to
nitrocellulose membranes in a semidry transfer cell. The transblotted
membrane was blocked with Tris-buffered saline/0.1% Tween 20
(TBST) containing 5% nonfat milk for 90 min, and the membrane
was incubated with the appropriate primary antibody in TBST/5%
nonfat milk at 4 °C overnight. After being washed three times with
TBST for a total of 45 min, the transblotted membrane was
incubated with goat antirabbit or antimouse IgG-horseradish per-
oxidase conjugates (diluted 1:1000) for 1 h at room temperature
and washed four times with TBST for a total of 1 h. The
immunoblots were visualized by enhanced chemiluminescence.
Semiquantitative PCR Analysis of mRNA Transcripts of
Cyclin D1 Gene. MCF-7 cells were subject to total RNA isolation
by using the RNeasy minikit (Qiagen, Valencia, CA). RNA
concentrations and quality were assessed spectrophotometrically
by measuring absorption at 260 nm. Aliquots of 20 µg of total RNA
from each sample were reverse-transcribed to cDNA using the
Omniscript RT kit (Qiagen) according to manufacturer’s instruc-
tions. The primers used were as follows: cyclin D1, forward, 5′-
ATGGAACACCAGCTCCTGTGCTGC-3′, reverse, 5′-TCAGAT-
GTCCACGTCCCGCACGT-3′, â-actin, forward, 5′-TCTACA-
ATGAGCTGCGTGTG-3′, reverse, 5′-GGTCAGGATCTTCAT-
GAGGT-3′. The reaction conditions were as follows: for cyclin
D1, (1) initial denaturation at 95 °C for 5 min, (2) 34 cycles of
amplification (95 °C for 1 min, 65 °C for 1 min 45 s, and 72 °C
for 1 min), and (3) a final extension step of 10 min at 72 °C; for
â-actin, (1) initial denaturation at 95 °C for 3 min, (2) 40 cycles of
amplification (95 °C for 30 s, 58 °C for 20 s, and 72 °C for 45 s),
and (3) a final extension step of 10 min at 72 °C. The PCR reaction
products were separated electrophoretically in a 1.2% agarose gel
and stained with ethidium bromide.
(1) Weinstein, I. B. Disorders in cell circuitry during multistage
carcinogenesis: the role of homeostasis. Carcinogenesis 2000, 21,
857-864.
(2) Chung, D. C. Cyclin D1 in human neuroendocrine: tumorigenesis.
Ann. N. Y. Acad. Sci. 2004, 1014, 209-217.
(3) Diehl, J. A. Cycling to cancer with cyclin D1. Cancer Biol. Ther.
2002, 1, 226-231.
(4) Stacey, D. W. Cyclin D1 serves as a cell cycle regulatory switch in
actively proliferating cells. Curr. Opin. Cell Biol. 2003, 15, 158-
163.
(5) Wang, C.; Li, Z.; Fu, M.; Bouras, T.; Pestell, R. G. Signal transduction
mediated by cyclin D1: from mitogens to cell proliferation:
a
molecular target with therapeutic potential. Cancer Treat. Res. 2004,
119, 217-237.
(6) Fu, M.; Wang, C.; Li, Z.; Sakamaki, T.; Pestell, R. G. Minireview.
Cyclin D1: normal and abnormal functions. Endocrinology 2004,
145, 5439-5447.
(7) Ewen, M. E.; Lamb, J. The activities of cyclin D1 that drive
tumorigenesis. Trends Mol. Med. 2004, 10, 158-162.
(8) Zhou, P.; Jiang, W.; Zhang, Y. J.; Kahn, S. M.; Schieren, I.; Santella,
R. M.; Weinstein, I. B. Antisense to cyclin D1 inhibits growth and
reverses the transformed phenotype of human esophageal cancer cells.
Oncogene 1995, 11, 571-580.
(9) Schrump, D. S.; Chen, A.; Consoli, U. Inhibition of lung cancer
proliferation by antisense cyclin D. Cancer Gene Ther. 1996, 3, 131-
135.
(10) Arber, N.; Doki, Y.; Han, E. K.; Sgambato, A.; Zhou, P.; Kim, N.
H.; Delohery, T.; Klein, M. G.; Holt, P. R.; Weinstein, I. B. Antisense
to cyclin D1 inhibits the growth and tumorigenicity of human colon
cancer cells. Cancer Res. 1997, 57, 1569-1574.
(11) Kornmann, M.; Arber, N.; Korc, M. Inhibition of basal and mitogen-
stimulated pancreatic cancer cell growth by cyclin D1 antisense is
associated with loss of tumorigenicity and potentiation of cytotoxicity
to cisplatinum. J. Clin. InVest. 1998, 101, 344-352.
(12) Saikawa, Y.; Kubota, T.; Maeda, S.; Otani, Y.; Kumai, K.; Kitajima,
M. Inhibition of DNA methyltransferase by antisense oligodeoxy-
nucleotide modifies cell characteristics in gastric cancer cell lines.
Oncol. Rep. 2004, 12, 527-531.
(13) Chen, B.; Zhang, X. Y.; Zhang, Y. J.; Zhou, P.; Gu, Y.; Fan, D. M.
Antisense to cyclin D1 reverses the transformed phenotype of human
gastric cancer cells. World J. Gastroenterol. 1999, 5, 18-21.
(14) Sauter, E. R.; Yeo, U. C.; von Stemm, A.; Zhu, W.; Litwin, S.;
Tichansky, D. S.; Pistritto, G.; Nesbit, M.; Pinkel, D.; Herlyn, M.;
Bastian, B. C. Cyclin D1 is a candidate oncogene in cutaneous
melanoma. Cancer Res. 2002, 62, 3200-3206.
Cell Viability Analysis. The effect of individual test agents on
cell viability was assessed by using the MTT assay in six replicates.
Cells were seeded and incubated in 96-well, flat-bottomed plates
in DMEM/Ham’s F-12 medium with 10% FBS for 24 h and were
exposed to various concentrations of test agents dissolved in DMSO
(final DMSO concentration, 0.1%) in 5% FBS-supplemented
DMEM/Ham’s F-12 medium. Controls received DMSO vehicle at
a concentration equal to that of drug-treated cells. The medium
was removed and replaced by 200 µL of 0.5 mM MTT in 10%
FBS-containing RPMI 1640 medium, and cells were incubated in
the 5% CO2 incubator at 37 °C for 2 h. Supernatants were removed
from the wells, and the reduced MTT dye was solubilized in 200
µL/well DMSO. Absorbance at 570 nm was determined on a plate
reader.
(15) Sauter, E. R.; Herlyn, M.; Liu, S. C.; Litwin, S.; Ridge, J. A.
Prolonged response to antisense cyclin D1 in a human squamous
cancer xenograft model. Clin. Cancer Res. 2000, 6, 654-660.
(16) Weinstein, I. B. Cancer. Addiction to oncogenes, the Achilles heel
of cancer. Science 2002, 297, 63-64.
(17) Huang, J. W.; Shiau, C. W.; Yang, Y. T.; Kulp, S. K.; Chen, K. F.;
Brueggemeier, R. W.; Shapiro, C. L.; Chen, C. S. Peroxisome
proliferator-activated receptor gamma-independent ablation of cyclin
D1 by thiazolidinediones and their derivatives in breast cancer cells.
Mol. Pharmacol. 2005, 67, 1342-1348.
Acknowledgment. This work is supported by National
Institutes of Health Grants CA-94829 and CA112250.
Supporting Information Available: Elemental analysis data.
This material is available free of charge via the Internet at http://
pubs.acs.org.
JM060057H