Thalidomide analogues demonstrate dual inhibition of both angiogenesis and prostate cancer

Article abstract of DOI:10.1016/j.bmc.2003.11.007

The identification of agents with antiproliferative activity against endothelial cells has significant value for the treatment of many angiogenesis-dependent pathologies. Herein, we describe the discovery of a series of thalidomide analogues possessing inhibitory effects against both endothelial and prostate cancer cells. More specifically, several analogues exhibited low micromolar to mid-nanomolar potency in the inhibition of human microvascular endothelial cell (HMEC) proliferation, both in the presence and absence of vascular endothelial growth factor (VEGF), with the tetrafluorophthalimido class of compounds demonstrating the greatest potency. Additionally, all the compounds were screened against two different androgen independent prostate cancer cell lines (PC-3 and DU-145). Again, the tetrafluorophthalimido analogues exhibited the greatest effect with GI ₅₀ values in the low micromolar range. Thalidomide was found to demonstrate selective inhibition of androgen receptor positive LNCaP prostate cancer cells. Furthermore, we showed that, as an example, tetrafluorophthalimido analogue 19 was able to completely inhibit the prostate specific antigen (PSA) secretion by the LNCaP cell line, while thalidomide demonstrated a 70% inhibition. We have also demonstrated that a correlation exists between HMEC and prostate cancer cell proliferation for this structural class. Altogether, our study suggests that these analogues may serve as promising leads for the development of agents that target both androgen dependent and independent prostate cancer and blood vessel growth.

Full text of DOI:10.1016/j.bmc.2003.11.007

Bioorganic & Medicinal Chemistry 12 (2004) 327–336
Thalidomide analogues demonstrate dual inhibition of both
angiogenesis and prostate cancer
Scott M. Capitosti, Todd P. Hansen and Milton L. Brown*
Department of Chemistry, University of Virginia, McCormick Road, PO Box 400319, Charlottesville, VA 22904, USA
Received 21 August 2003; revised 28 October 2003; accepted 10 November 2003
Abstract—The identification of agents with antiproliferative activity against endothelial cells has significant value for the treatment
of many angiogenesis-dependent pathologies. Herein, we describe the discovery of a series of thalidomide analogues possessing
inhibitory effects against both endothelial and prostate cancer cells. More specifically, several analogues exhibited low micromolar
to mid-nanomolar potency in the inhibition of human microvascular endothelial cell (HMEC) proliferation, both in the presence
and absence of vascular endothelial growth factor (VEGF), with the tetrafluorophthalimido class of compounds demonstrating the
greatest potency. Additionally, all the compounds were screened against two different androgen independent prostate cancer cell
lines (PC-3 and DU-145). Again, the tetrafluorophthalimido analogues exhibited the greatest effect with GI50 values in the low
micromolar range. Thalidomide was found to demonstrate selective inhibition of androgen receptor positive LNCaP prostate can-
cer cells. Furthermore, we showed that, as an example, tetrafluorophthalimido analogue 19 was able to completely inhibit the
prostate specific antigen (PSA) secretion by the LNCaP cell line, while thalidomide demonstrated a 70% inhibition. We have also
demonstrated that a correlation exists between HMEC and prostate cancer cell proliferation for this structural class. Altogether,
our study suggests that these analogues may serve as promising leads for the development of agents that target both androgen
dependent and independent prostate cancer and blood vessel growth.
# 2003 Elsevier Ltd. All rights reserved.
1. Introduction
tive effects in humans, an association was reported of
Angiogenesis is the process of new blood vessel growth,
and involves the proliferation of endothelial cells in
response to specific growth stimuli such as vascular
endothelial growth factor (VEGF) or basic fibroblast
growth factor (bFGF). Angiogenesis is an important
natural process occurring in the body, both in health
(wound healing,1 female reproductive cycle2) and dis-
ease (cancer,3 rheumatoid arthritis,4 diabetic retino-
pathy5). Vessel growth is controlled by a balance of
endogenous inhibitors and stimulators. The growth and
maintenance of solid tumors is highly dependent on
neovascularization and can be regulated by compounds
that interfere with either the stimulation or proliferation
of endothelial cells.6 As a result, the control of angio-
genesis continues to be an attractive area for novel
therapeutic agent development.3
teratogenic limb defects from maternal thalidomide
usage.7 Aside from this serious teratogenic effect on the
fetus, the drug does have therapeutic value: (1) for its
immunosuppressive effect in the treatment of graft ver-
sus host disease;8,9 (2) in the treatment of leprosy;10,11
and (3) for inflammatory dermatoses.12 In addition,
thalidomide has significant anti-angiogenic activity, and
the effects of thalidomide on corneal angiogenesis
induced by vascular endothelial growth factor (VEGF)
have been reported.13 Additionally, thalidomide has
demonstrated inhibitory effects on angiogenesis in the
basic fibroblast growth factor (bFGF) induced rabbit
corneal micropocket assay14 and orally in mice
models.15
Prostate cancer is the most common malignancy in
American men and is the second leading cause of cancer
mortality.16 Prostate cancer, as is the case with numer-
ous types of cancer, is dependent on the recruitment of
new blood vessels to grow and metastasize. In fact, in
prostate cancer, an increased microvessel density corre-
lates to poorer prognosis.17 As long as the cancer is
confined to the prostate, it can be successfully con-
trolled by radiation or surgery. However, in metastatic
disease, few treatment options are available beyond
One such agent is thalidomide (1) (Fig. 1), which was
developed in the 19500s by Chemie Grunenthal of Ger-
many as a non-toxic sedative.7 In addition to its seda-
Keywords: Phthalimide; Prostate cancer; Structure–activity relation-
ships; Endothelial cells.
* Corresponding author. Tel.: +1-434-982-3091; fax: +1-434-924-
0798; e-mail: mlb2v@virginia.edu
0968-0896/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2003.11.007

328
S. M. Capitosti et al. / Bioorg. Med. Chem. 12 (2004) 327–336
androgen ablation.18 Clinical trials of thalidomide in
patients with androgen-independent prostate cancer have
been reported, and they show thalidomide has modest
activity in patients with metastatic prostate cancer.19
appropriate phthalic anhydride with an appropriate
aniline or primary amine derivative in refluxing acetic
acid.27 Compounds 2 and 20 were purchased from
Aldrich Chemical Company. Moderate to high yields of
34–81% were obtained without optimization of the
reaction conditions (Table 1).
Inhibitors of angiogenesis can exert their effect in many
ways, depending on their mechanism of action. Some
inhibit endothelial cells directly, while others inhibit the
angiogenesis signaling cascade or block the ability of
endothelial cells to break down the extracellular matrix.
Therapeutic agents that can inhibit more than one of
these pathways could prove to have a distinct advantage
over conventional anti-angiogenic therapies in which only
one mechanism of action is targeted. On the basis of the
latter statement, we have been engaged in the structural
modification of thalidomide with the aim of developing
novel anti-angiogenic/anti-cancer analogues that possess
superior biological action relative to thalidomide.
The second series of compounds including restricted
analogue targets 22–27 in Scheme 2 were also generated
from the condensation of an amine derivative with a
phthalic anhydride derivative in refluxing acetic acid.
Yields of 25–90% were obtained for these analogues
(Table 2). Amine 29, used in the preparation of 26 and
27, was realized by conversion of the ketone precursor, 1-
benzosuberone, to the oxime 28 utilizing the procedure
of Ballini et al.,28 followed by catalytic hydrogenation of
28 with 10% Pd/C and hydrogen gas (Scheme 3).
Herein we report that substitution of the four aromatic
hydrogen atoms of phthalimide analogues of thalido-
mide with fluorine leads to more potent angiogenic and
prostate cancer cell inhibitory activity (Fig. 2). The tet-
rafluorophthalimides were synthesized for the following
reasons. First, Ng et al. recently demonstrated that tet-
rafluorinated thalidomide analogues are inhibitors of
angiogenesis,20 and secondly, it has been reported that
the glutarimide ring of thalidomide is not essential for
antimetastatic activity against B16BL6 melanoma cells,
since a completely hydrolyzed glutarimide group did not
render the molecule inactive.21 That evidence, coupled
with the numerous published examples of various N-
phenylphthalimides that possess increased TNF-a reg-
ulatory activity over thalidomide,22ꢀ26 lends support to
the notion that structural simplification of the glutarimide
moiety does not produce biologically inactive analogues.
3. Biological activity studies
Thalidomide analogues 2–27 were first screened for
their ability to inhibit the proliferation of human
microvascular endothelial cells (HMEC’s), both in the
presence and absence of vascular endothelial growth
factor (VEGF), with thalidomide as the standard. It has
been demonstrated previously that inhibition of angio-
genesis by thalidomide requires metabolic activation;29
however, unactivated thalidomide was used as a stan-
dard in our proliferation assays to simply demonstrate
the ability of our analogues to inhibit angiogenesis
without the need for prior metabolic activation. Addition-
ally, the above-mentioned analogues were screened for
inhibitory activity against the proliferation of PC-3 and
DU-145 prostate cancer cell lines. The results obtained for
both of the above experiments are shown in Table 3.
In the non-fluoro-substituted phthalimide class (2–10,
20, 22, 24, 26), the most active compound against
HMEC proliferation (6) exhibited low micromolar IC50
values both in the presence and absence of VEGF.
Mechanistically interesting was the fact that analogue 6
demonstrated potent ability to overcome growth factor
mediated proliferation of endothelial cells. Altogether,
2. Results and discussion
2.1. Chemistry
The first series of target compounds 3–19 and 21 in
Scheme 1 were synthesized by the condensation of an
Figure 1. Thalidomide (1).
Scheme 1. Synthesis of target phthalimide analogues. Reagents: (a)
acetic acid, reflux, 2–3 h.
Scheme 2. Synthesis of target phthalimide restricted analogues.
Reagents: (a) acetic acid, reflux, 2–3 h.
Figure 2. Design strategy of first generation analogues.

S. M. Capitosti et al. / Bioorg. Med. Chem. 12 (2004) 327–336
329
Table 1. Target phthalimides in series 1
Table 3. Inhibitory effects of compounds 2–27 on HMEC and pros-
tate cancer cell proliferationa
Compd
R
Compd
HMECb IC50 (mM)
Prostate cancer GI50 (mM)c
1
2
3
4
R’
n
Yield (%)
(+) VEGF (ꢀ) VEGF
PC-3
DU-145
a
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
H
H
H
H
H
H
H
H
H
F
F
F
F
F
F
F
F
F
H
H
H
H
H
H
H
H
H
F
F
F
F
F
F
F
F
F
H
H
H
H
H
H
H
H
H
F
F
F
F
F
F
F
F
F
H
H
H
H
H
H
H
H
H
F
F
F
F
F
F
F
F
F
H
4-Cl
3-Cl
2-Cl
3,4-Cl
4-CH3
3-CH3
2-CH3
4-OCH3
H
4-Cl
3-Cl
2-Cl
3,4-Cl
4-CH3
3-CH3
2-CH3
4-OCH3
H
0
—
070
064
050
067
077
081
069
071
0
045
051
035
057
075
078
034
065
1
Thalidomide
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
>300
>300
>300
>300
304ꢁ16.7
>300
>300
>300
>300
>300
>300
321ꢁ178
4.91
>300
>300
74.4ꢁ27.8
>300
>300
>300
>300
>300
>300
>300
11.8
302ꢁ21.5 242ꢁ29.0
71.1ꢁ12.8 51.6ꢁ5.81
181ꢁ11.1 155ꢁ16.9
>300
>300
10.6ꢁ6.2015.0 ꢁ4.11
71.6ꢁ31.9 68.9ꢁ35.8
167ꢁ14.6 200ꢁ4.19
187ꢁ3.54 157ꢁ9.84
208ꢁ23.6 144ꢁ27.6
0.77ꢁ0.23 0.68ꢁ0.11
5.98ꢁ2.65 1.78ꢁ0.26
0.78ꢁ0.05 0.42ꢁ0.19
8.37ꢁ2.24 7.09ꢁ0.83
70
10.7ꢁ3.83
8.33ꢁ1.98
0.77
3.42
2.03
3.07
>300d
<0.1
<0.1
>300d
0.33ꢁ0.18
<0.1
4.98
17.0
1.16
1.77
2.12
>300
0.88
>300
1.01
11.6
12.2
5.34
7.57
3.97
>300
1.97
a
H
F
H
F
H
F
H
F
—
65
2.40ꢁ0.21 1.73ꢁ0.36
2.28ꢁ0.54 1.53ꢁ0.25
262ꢁ35.3 62.2ꢁ26.3
0.95ꢁ0.16 0.91ꢁ0.21
H
1
a Commercially available from Aldrich Chemical Co.
>300
147ꢁ50.1
385ꢁ225
1.07ꢁ0.01 2.40ꢁ0.42
258ꢁ17.2 306ꢁ13.9
0.19ꢁ0.07 0.18ꢁ0.02
251ꢁ87.9 28.4ꢁ2.73
0.62ꢁ0.08 0.32ꢁ0.04
0.91
116ꢁ21.1
1.53
218ꢁ19.7
Table 2. Target phthalimides in series 2
1.88
>300
0.55
>300
1.48
Compd
R
a All experiments were run in triplicate, and theꢁvalues represent the
b Human microvessel endothelial cells.
1
2
3
4
R0
n
Yield (%)
SEM.
22
23
24
25
26
27
H
F
H
F
H
F
H
F
H
F
H
F
H
F
H
F
H
F
H
F
H
F
H
F
H
H
H
H
H
H
1
1
2
2
3
3
90
77
84
53
78
25
c Prostate cancer GI50 values for compounds 11–19, 21, 23, 25, and 27
were determined by the National Cancer Institute Developmental
Therapeutics Program and all others were determined in our labora-
tory.
d The poor solubility of this compound in the biological medium may
have contributed to the results obtained.
gues, as they both revealed a remarkable increase in
potency over thalidomide in the HMEC screen with
IC50 values ranging from 100–330 nm (Table 3). Fur-
thermore, all tetrafluoro-substituted analogues pos-
sessed a considerable increase in potency in anti-
angiogenic activity compared to thalidomide. Similarly,
in the prostate cancer screen, all tetrafluoro-analogues
demonstrated significant increases in potency over tha-
lidomide in both PC-3 and DU-145 cell lines.
Scheme 3. Synthesis of amine 29. Reagents: (a) NH2OH-HCl, ethanol,
amberlyst A-21, 24 h, rt (82%); (b) (i) 10% Pd/C, H2; (g), ethanol, 3
days; (ii) Et2O/HCl, (100%).
when compared to thalidomide, the non-fluoro sub-
stituted analogues (with the exception of 3, 6, and 7)
possessed either comparable or only slightly more potent
anti-angiogenic activity with and without VEGF present.
Compounds 16 and 17 possess para and meta methyl
groups on the phenyl ring attached to the imide nitro-
gen respectively (Table 1). Replacement of the para
methyl group of 16 with a para chloro group (12) results
in a slight loss in activity, raising the IC50 to the low
micromolar range. Replacement of the meta methyl
group of 17 with a meta chloro group (13) results in
only a modest decrease in activity to the high nanomo-
lar range. Both methyl and chloro groups add similar
contributions to the overall lipophilicity of the com-
pound; therefore, the electron donating properties of the
methyl group may play a more critical role in the
increased activity of compounds 16 and 17. The unsub-
stituted tetrafluoro-analogue (11) shows activity similar
to that of compounds 12 and 13, lending more support
to the electrostatic effect of the methyl group. Com-
Against prostate cancer cell lines, only compound 3
showed any appreciable increase in potency over thali-
domide in the prostate cancer screen, and this effect was
only seen with the DU-145 cell line. These results sug-
gest that there is no clear advantage gained by the sub-
stitution of a phenyl ring within this class. However, this
structural class does have efficacy against growth factor
mediated microvessel endothelial cell proliferation.
In the tetrafluorophthalimido class (11–19, 21, 23, 25,
27), compounds 16 and 17 were the most active analo-

330
S. M. Capitosti et al. / Bioorg. Med. Chem. 12 (2004) 327–336
pound 19, incorporating a para methoxy group, also has
similar activity to 12 and 13, possibly due to the meth-
oxy groups moderately inductive electron withdrawing
properties.
for the LNCaP subtype, or (3) androgen receptor posi-
tive prostate cancer has a metabolic profile that is
different from androgen receptor negative prostate can-
cer. Further studies should be warranted in these areas.
Tetrafluoro-analogue 21, with a methylene insertion
between the phenyl group and the imide nitrogen, also
shows significant anti-angiogenic and anti-prostate cancer
activity in the high nanomolar to low micromolar range.
Additionally, the tetrafluoro-substituted restricted analo-
gues 23, 25, and 27 show similar to slightly increased anti-
angiogenic and anti-prostate cancer activity relative to
both 21 and the tetrahydro counterparts 22, 25, and 26.
Consequently, it was found that restricting the rotation of
the phenyl ring in 21 does not prove to be detrimental to
the compounds in vitro potency (Figs 3 and 4).
We have also found that PSA secretion can be inhibited
by thalidomide treatment on a per cell basis. Thalido-
mide showed a 70% inhibition in the amount of PSA
per cell at its IC50 concentration for the androgen recep-
tor positive LNCaP cell line (Table 4). We demonstrate,
as an example, that the tetrafluorophthalimido analogue
19 showed a 100% inhibition of PSA secretion at its IC50
concentration. Altogether, the prostate cancer inhibi-
tion profile for analogue 19 is considerably better than
thalidomide with demonstrated effectiveness against
both androgen positive and negative prostate cancer
and PSA secretion.
In order to broaden the activity spectrum of the tetra-
fluorophthalimido class, we selected one tetrafluoro
analogue to study the effect against the androgen
dependent LNCaP prostate cancer cell line. Analogue
19 was chosen for this study because of its consistent
potency profile across all four biological experiments
presented in Table 3. As shown in Table 4, analogue 19,
with an IC50 of 15.6 mM, proved to be of equal potency
as thalidomide at inhibiting the proliferation of the
LNCaP cells. Further, this data confirms that thalido-
mide has selective activity towards androgen receptor
positive prostate cancer (LNCaP IC50 25.3ꢁ21.1 mM,
Table 4) in comparison to androgen receptor negative
prostate cancer (IC50>300 mM for DU-145 and PC-3,
Table 3). This finding is different from studies on
angiogenesis where thalidomide must be metabolically
activated.29 Our data suggests that thalidomide will be
(1) directly effective in androgen receptor positive pros-
tate cancer cell lines, (2) does not need to be activated
Unfortunately for the prostate cancer patient, a poor
prognosis has been correlated with increased micro-
vessel density.17 We have found a significant correlation
exists between HMEC and PC-3 proliferation (Plots A
and B in Fig. 3), even in the presence of the vascular
growth factor. This is significant and suggests that our
compounds have ability to overcome this growth sti-
Table 4. Effect of 19 on LNCaP proliferation and PSA secretiona
Compd
LNCaP IC50 (mM)
%PSA/cellb
Thalidomide
19
25.3ꢁ21.1
15.6ꢁ1.37
30
0
a All experiments were run in triplicate, and the ꢁvalues represent the
SEM.
b %PSA per cell was measured relative to control at the IC50 dose of
drug. Controls were with (0.2%) DMF vehicle.
Figure 3. Correlation plots of (A): log [PC-3 IC50] versus log [(ꢀ)VEGF IC50]; and (B): log [PC-3 IC50] versus log [(+)VEGF IC50].
Figure 4. Correlation plots of (A): log [DU-145 IC50] versus log [(ꢀ)VEGF IC50]; and (B): log [DU-145 IC50] versus log [(+)VEGF IC50].

S. M. Capitosti et al. / Bioorg. Med. Chem. 12 (2004) 327–336
331
mulus pathway. In less aggressive DU-145 cells, our
5.2. Analytical procedures
compounds demonstrated a lesser ability to inhibit in
the presence of VEGF (Plot B in Fig. 4). This suggests
that our compounds have a much greater efficacy
against the more aggressive androgen independent
prostate cancer phenotype.
Melting points were determined in open capillary tubes
on an Electrothermal melting point apparatus and are
1
uncorrected. H and 13C NMR spectra were measured
at 25 ꢂC on compounds in solution in CDCl3 or DMSO-
d6 on a GE 300 MHz spectrometer. APCI mass spectra
were obtained on a Finnagan LcQ Classic. The Uni-
versity of Illinois at Urbana-Champaign Mass Spectro-
metry Laboratory performed the EI analyses. IR
spectra were recorded on a Nicolet Impact 400D spec-
trophotometer. Elemental analyses were performed by
Atlantic Microlab, and microanalytical data were with-
inꢁ0.4% of the calculated figures. Thin-layer chroma-
tography was done on precoated aluminum silica gel
plates (silica gel 60F 254).
4. Conclusion
Our study has identified new thalidomide analogues
with dual anti-angiogenic/anti-prostate cancer activity
(analogues 16, 17, 21, 23, 25, and 27). The preliminary
SAR with these compounds revealed the importance of
tetrafluorination of the phthalimide core, in addition to
para or meta methyl substitution on the opposite phenyl
ring. Furthermore, methylene insertion between the
phenyl ring and the imide nitrogen in the tetra-
fluorophthalimido class does not abrogate in vitro
activity, nor does restricting the rotation of the freely
rotatable phenyl group.
5.3. General procedure for preparation of phthalimide
derivatives
The appropriate phthalic anhydride and amine were
stirred under reflux in acetic acid for 2–4 h. The solution
was poured into water, and the resulting precipitate fil-
tered off by suction and recrystallized from absolute
ethanol.
The presence of the fluorine substituents on the phtha-
limide core drastically increases the biological potency
of the aforementioned analogues. Since the fluorine
atom is a close steric mimic of hydrogen; incorporation
of fluorine into drugs often improves their biological
potency.30 Such improvements have been attributed to
either the enhanced lipophilicity of fluorinated com-
pounds or to stereoelectronic changes imparted to
compounds by the strongly electronegative fluorine
substitution resulting in more effective binding to their
respective target(s). Additionally, the fluorinated com-
pounds presented herein represent analogues that would
be resistant to the phenyl metabolism that has been
reported for thalidomide.31
5.3.1. N-(4-Chlorophenyl)-phthalimide (3). Compound 3
was prepared from phthalic anhydride (5.0g/33.8
mmol) and 4-chloroaniline (4.31 g/33.8 mmol): mp 189–
192 ꢂC (lit.27 193–194 ꢂC); H NMR (CDCl3) d 7.4–7.5
1
(m, 4H, ArH), 7.7–8.0(m, 4H, Ar H); 13C NMR
(CDCl3) d 167.5, 135.1, 134.3, 132.1, 130.7, 129.8, 128.2,
124.4; APCI m/z (rel intensity) 258.3 (M+, 100).
5.3.2. N-(3-Chlorophenyl)-phthalimide (4). Compound 4
was prepared from phthalic anhydride (5.0g/33.8
mmol) and 3-chloroaniline (4.31 g/33.8 mmol/3.58 mL):
mp 160–163 ꢂC (lit.32 164 ꢂC); H NMR (CDCl3) d 7.3–
1
The evidence presented herein indicates the dramatically
increased activity and dual acting nature of the tetra-
fluorophthalimido analogues, and therefore offers a
unique opportunity for utilizing these fluorinated ana-
logues as leads in the search for novel therapeutic agents
for prostate cancer. These compounds demonstrate
increased effectiveness against prostate cancer in com-
parison to thalidomide and show ability to inhibit both
phenotypes. In particular, the fluorinated analogues
seem to be more active against the androgen indepen-
dent PC-3/DU-145 cells, and it is precisely this andro-
gen independent prostate cancer that proves to be more
fatal to patients because of a current lack of effective
treatment strategies. Further, the use of these com-
pounds could be extended towards evaluating them
towards other cancers in which poor patient prognosis
correlates to increased microvessel density.
7.5 (m, 4H, ArH), 7.7–8.0(m, 4H, Ar H); 13C NMR
(CDCl3) d 167.3, 135.1, 133.3, 132.0, 130.5, 128.7, 127.1,
125.1, 124.4; APCI m/z (rel intensity) 258.3 (M+, 100).
5.3.3. N-(2-Chlorophenyl)-phthalimide (5). Compound 5
was prepared from phthalic anhydride (300 mg/2.03
mmol) and 2-chloroaniline (260 mg/2.03 mmol/0.21
mL): mp 143–145 ꢂC (lit.33 143 ꢂC); 1H NMR (CDCl3) d
7.2–8.0(m, 8H, Ar H); 13C NMR (CDCl3) d 166.2,
134.1, 132.8, 131.5, 130.3, 130.1, 129.2, 127.4, 123.6,
123.2; APCI m/z (rel intensity) 256.7 (100), 258.6 (M+,
60).
5.3.4. N-(3,4-Dichlorophenyl)-phthalimide (6). Com-
pound 6 was prepared from phthalic anhydride (920
mg/6.2 mmol) and 3,4-dichloroaniline (1.0g/6.2 mmol):
mp 196–198 ꢂC (lit.34 198 ꢂC); H NMR (CDCl3) d 7.3–
1
7.6 (m, 3H, ArH), 7.7–8.0(m, 4H, Ar H); 13C NMR
(CDCl3) d 167.1, 135.3, 131.9, 131.2, 128.7, 126.0, 124.5;
APCI m/z (rel intensity) 292.2 (M+, 100), 294.2 (65).
5. Experimental
5.1. Materials
5.3.5. N-(4-Methylphenyl)-phthalimide (7). Compound 7
was prepared from phthalic anhydride (1.4 g/9.3 mmol)
and p-toluidine (1.0 g/9.3 mmol/1.03 mL): mp 204–
Reagents were purchased from Aldrich or Lancaster
and were used as received. Solvents used were pre-
viously dried by distillation.
206 ꢂC (lit.35 201–203 ꢂC); H NMR (CDCl3) d 2.4 (s,
1

332
S. M. Capitosti et al. / Bioorg. Med. Chem. 12 (2004) 327–336
3H, CH3), 7.3 (s, 4H, ArH), 7.7–8.0(m, 4H, Ar H); 13C
NMR (CDCl3) d 167.9, 138.7, 134.8, 132.4, 130.3, 129.5,
127.0, 124.2, 21.7; APCI m/z (rel intensity) 238.2 (M+,
100).
5.3.12. N-(2-Chlorophenyl)-tetrafluorophthalimide (14).
Compound 14 was prepared from tetrafluorophthalic
anhydride (400 mg/1.8 mmol) and 2-chloroaniline (230
mg/1.8 mmol/0.19 mL): mp 161–163 ꢂC; 1H NMR
(DMSO-d6) d 7.5–7.7 (m, 3H, ArH), 7.7–7.8 (m, 1H,
ArH); 13C NMR (DMSO-d6) d 161.1, 146.3, 144.7,
5.3.6. N-(3-Methylphenyl)-phthalimide (8). Compound 8
was prepared from phthalic anhydride (300 mg/2.03
mmol) and m-toluidine (220 mg/2.03 mmol/0.22 mL):
132.1, 131.8, 131.3, 130.2, 128.5, 128.2; EI m/z (rel
+
intensity) 329.0(M
,
10), 294.0 (100). Anal.
mp 177–180 ꢂC (lit.36 175–178 ꢂC); H NMR (CDCl3) d
(C14H4ClF4NO2) C, H, N; C: calcd, 51.01; found, 51.02;
H: calcd, 1.22; found, 1.10; N: calcd, 4.25; found, 4.20.
1
2.4 (s, 3H, CH3), 7.2–7.5 (m, 4H, ArH), 7.8–8.0(m, 4H,
ArH); 13C NMR (CDCl3) d 167.0, 138.8, 134.0, 131.4,
131.1, 128.7, 128.6, 126.9, 123.4, 21.0; APCI m/z (rel
intensity) 237.6 (M+, 100).
5.3.13.
N-(3,4-Dichlorophenyl)-tetrafluorophthalimide
(15). Compound 15 was prepared from tetra-
fluorophthalic anhydride (200 mg/0.91 mmol) and 3,4-
dichloroaniline (150 mg/0.91 mmol): mp 206–209 ꢂC; 1H
NMR (DMSO-d6) d 7.4 (m, 1H, ArH), 7.7–7.9 (m, 2H,
ArH); 13C NMR (DMSO-d6) d 161.4, 146.1, 144.5,
142.6, 141.0, 131.6, 131.4, 131.2, 130.8, 129.2, 127.7, 113.9;
EI m/z (rel intensity) 362.9 (M+, 100), 364.4 (68). Anal.
(C14H3Cl2F4NO2) C, H, N; C: calcd, 46.19; found, 45.98;
H: calcd, 0.83; found, 0.96; N: calcd, 3.85; found, 3.85.
5.3.7. N-(2-Methylphenyl)-phthalimide (9). Compound 9
was prepared from phthalic anhydride (300 mg/2.03
mmol) and o-toluidine (220 mg/2.03 mmol/0.22 mL):
mp 184–186 ꢂC (lit.37 182–183 ꢂC); H NMR (CDCl3) d
1
2.3 (s, 3H, CH3), 7.2–7.5 (m, 4H, ArH), 7.7–8.0(m, 4H,
ArH); 13C NMR (CDCl3) d 167.0, 136.2, 134.0, 131.6,
130.8, 129.1, 128.4, 126.5, 126.2, 123.4, 17.7; APCI m/z
(rel intensity) 237.7 (M+, 100).
5.3.14. N-(4-Methylphenyl)-tetrafluorophthalimide (16).
Compound 16 was prepared from tetrafluorophthalic
anhydride (200 mg/0.91 mmol) and p-toluidine (100 mg/
0.91 mmol/0.1 mL): mp 234–236 C; H NMR (CDCl3)
d 2.4 (s, 3H, CH3), 7.2–7.4 (m, 4H, ArH); 13C NMR
(CDCl3) d 161.2, 146.5, 145.0, 143.0, 138.7, 129.6, 127.4,
125.9, 20.9; EI m/z (rel intensity) 309.0 (M+, 100).
Anal. (C15H7F4NO2) C, H, N; C: calcd, 58.26; found,
58.48; H: calcd, 2.28; found, 2.31; N: calcd, 4.53;
found, 4.58.
5.3.8. N-(4-Methoxyphenyl)-phthalimide (10). Com-
pound 10 was prepared from phthalic anhydride (1.2 g/
8.1 mmol) and p-anisidine (1.0g/8.1 mmol): mp 162–
ꢂ
1
164 ꢂC (lit.38 162 ꢂC); H NMR (CDCl3) d 3.8 (s, 3H,
1
CH3), 7.0(d, J=8.85 Hz, 2H, ArH), 7.3 (d, J=8.86 Hz,
2H, ArH), 7.7–8.0(m, 4H, Ar H); 13C NMR (CDCl3) d
168.1, 159.8, 134.8, 132.3, 128.4, 124.8, 124.2, 115.0,
56.0; APCI m/z (rel intensity) 254.2 (M+, 100).
5.3.9. N-Phenyl-tetrafluorophthalimide (11). Compound
11 was prepared from tetrafluorophthalic anhydride
(200 mg/0.91 mmol) and anilinꢂe (80 mg/0.91 mmol/0.08
mL): mp 206–208 ꢂC (lit.39 202 C); 1H NMR (CDCl3) d
7.3–7.6 (m, 5H, ArH); 13C NMR (CDCl3) d 161.0,
146.2, 145.0, 143.1, 141.6, 130.1, 129.0, 128.5, 126.1,
113.9; APCI m/z (rel intensity) 296.2 (M+, 100).
5.3.15. N-(3-Methylphenyl)-tetrafluorophthalimide (17).
Compound 17 was prepared from tetrafluorophthalic
anhydride (400mg/1.8 mmol) and m-toluidine (190mg/
1.8 mmol/0.20 mL): mp 188–191 ꢂC; 1H NMR (DMSO-
d6) d 2.3 (s, 3H, CH3), 7.2–7.3 (m, 3H, ArH), 7.4–7.5 (t,
J=7.45, 7.66 Hz, 1H, ArH); 13C NMR (DMSO-d6) d
161.9, 143.1, 140.9, 138.7, 130.9, 129.5, 129.0, 127.8,
124.5, 20.9; EI m/z (rel intensity) 309.1 (M+, 100). Anal.
(C15H7F4NO2) C, H, N; C: calcd, 58.26; found, 58.08;
H: calcd, 2.28; found, 2.27; N: calcd, 4.53; found, 4.59.
5.3.10. N-(4-Chlorophenyl)-tetrafluorophthalimide (12).
Compound 12 was prepared from tetrafluorophthalic
anhydride (200 mg/0.91 mmol) and 4-chloroaniline (120
mg/0.91 mmol): mp 260–263 ꢂC; H NMR (DMSO-d6)
1
d 7.4 (d, J=8.67 Hz, 2H ArH), 7.6 (d, J=8.66 Hz, 2H,
5.3.16. N-(2-Methylphenyl)-tetrafluorophthalimide (18).
Compound 18 was prepared from tetrafluorophthalic
anhydride (400mg/1.8 mmol) and o-toluidine (190mg/
1.8 mmol/0.19 mL): mp 143–145 ꢂC; 1H NMR (DMSO-
d6) d 2.1 (s, 3H, CH3), 7.3–7.4 (m, 4H, ArH); 13C NMR
(DMSO-d6) d 161.8, 145.9, 144.5, 136.5, 130.9, 129.9,
129.7, 129.0, 126.8, 17.3; EI m/z (rel intensity) 309.1
(M+, 100). Anal. (C15H7F4NO2) C, H, N; C: calcd,
58.26; found, 58.54; H: calcd, 2.28; found, 2.17; N:
calcd, 4.53; found, 4.56.
ArH); 13C NMR (DMSO-d6) d 161.7, 146.1, 144.3,
142.5, 140.9, 133.3, 129.8, 129.2, 114.0; EI m/z (rel
+
intensity) 329.0(M
,
100), 331.2 (36). Anal.
(C14H4ClF4NO2) C, H, N; C: calcd, 51.01; found,
50.81; H: calcd, 1.22; found, 1.44; N: calcd, 4.25;
found, 4.18.
5.3.11. N-(3-Chlorophenyl)-tetrafluorophthalimide (13).
Compound 13 was prepared from tetrafluorophthalic
anhydride (400 mg/1.8 mmol) and 3-chloroaniline (230
mg/1.8 mmol/0.19 mL): mp 182–184 ꢂC; 1H NMR
(DMSO-d6) d 7.4–7.5 (m, 2H, ArH), 7.6–7.7 (m, 2H,
ArH); 13C NMR (DMSO-d6) d 161.6, 146.0, 144.4,
142.6, 133.1, 132.3, 130.8, 128.8, 127.3, 126.2; EI m/z
(rel intensity) 329.0(M +, 100), 331.2 (35). Anal.
(C14H4ClF4NO2) C, H, N; C: calcd, 51.01; found,
50.99; H: calcd, 1.22; found, 1.08; N: calcd, 4.25;
found, 4.27.
5.3.17. N-(4-Methoxyphenyl)-tetrafluorophthalimide
(19). Compound 19 was prepared from tetra-
fluorophthalic anhydride (200 mg/0.91 mmol) and p-
ꢂC; 1H
anisidine (110mg/.091 mmol): mp 215–218
NMR (CDCl3) d 3.8 (s, 3H, CH3), 7.0(d, J=9.05 Hz,
2H, ArH), 7.3 (d, J=9.05 Hz, 2H, ArH); 13C NMR
(CDCl3) d 161.3, 159.4, 146.5, 145.1, 145.0, 143.0, 141.5,
141.4, 127.4, 122.6, 114.3, 55.2; EI m/z (rel intensity)

S. M. Capitosti et al. / Bioorg. Med. Chem. 12 (2004) 327–336
333
325.2 (M+, 100). Anal. (C15H7F4NO3) C, H, N; C:
calcd, 55.40; found, 55.52; H: calcd, 2.17; found, 2.20;
N: calcd, 4.31; found, 4.33.
126.1, 125.8, 49.5, 28.8, 27.2, 21.7; EI m/z (rel intensity)
349.0(M +, 5), 130.0 (100). Anal. (C18H11F4NO2) C, H,
N; C: calcd, 61.90; found, 61.65; H: calcd, 3.17; found,
3.16; N: calcd, 4.01; found, 3.97.
5.3.18. N-Benzyltetrafluorophthalimide (21). Compound
21 was prepared from tetrafluorophthalic anhydride
(400mg/1.8 mmol) and benzylamine (190mg/1.8 mmol/
5.3.23.
2-(6,7,8,9-Tetrahydro-5H-benzocyclohepten-5-
yl)-isoindole-1,3-dione (26). Compound 26 was from
phthalic anhydride (250mg/1.69 mmol) and amine 29
(272 mg/1.69 mmol): mp 196–198 ꢂC; 1H NMR
(DMSO-d6) d 1.2–1.3 (m, 1H), 1.5–2.0(m, 5H), 2.8–2.9
(bs, 2H), 5.1–5.2 (t, J=8.47, 8.67 Hz, 1H, NCH), 7.1–
7.8 (m, 7H, ArH), 8.8 (d, J=7.32 Hz, 1H, ArH); 13C
NMR (DMSO-d6) d 168.9, 168.4, 143.7, 141.6, 139.7,
132.1, 130.1, 130.0, 128.8, 127.2, 126.8, 125.7, 53.1, 36.0,
35.4, 29.9, 28.0; APCI m/z (rel intensity) 292.4 (M+,
100). Anal. (C19H17NO2–1H2O) C, H, N; C: calcd,
73.77; found, 73.80; H: calcd, 6.19; found, 6.17; N:
calcd, 4.53; found, 4.57.
0.20 mL): mp 179–181 ꢂC; H NMR (CDCl3) d 4.8 (s,
1
2H, CH2), 7.3–7.5 (m, 5H, ArH); 13C NMR (CDCl3) d
161.8, 146.2, 144.9, 134.8, 128.5, 128.4, 128.0, 42.0; EI
m/z (rel intensity) 309.1 (M+, 100). Anal.
(C15H7F4NO2) C, H, N; C: calcd, 58.26; found, 58.19;
H: calcd, 2.28; found, 2.20; N: calcd, 4.53; found, 4.51.
5.3.19. 2-Indan-1-yl-isoindole-1,3-dione (22). Compound
22 was prepared from phthalic anhydride (100 mg/0.675
mmol) and 1-aminoindan (90 mg/0.675 mmol/0.09 mL):
mp 173–176 ꢂC; 1H NMR (DMSO-d6) d 1.8–1.9 (m, 1H,
NCHCH2CH2), 2.3–2.4 (m, 1H, NCHCH2CH2), 2.7–
2.9 (m, 2H, NCHCH2CH2), 5.4–5.5 (q, J=8.09 Hz, 1H,
NCH), 7.1–7.7 (m, 7H, ArH), 8.6 (d, J=8.47 Hz, 1H,
ArH); 13C NMR (DMSO-d6) d 169.4, 168.9, 145.1,
143.7, 139.6, 132.1, 131.6, 130.1, 130.0, 128.7, 128.2,
127.2, 125.3, 54.9, 33.4, 30.7; APCI m/z (rel intensity)
263.8 (M+, 100). Anal. (C17H13NO2–1H2O) C, H, N; C:
calcd, 72.58; found, 72.50; H: calcd, 5.37; found, 5.39;
N: calcd, 4.98; found, 5.00.
5.3.24. 4,5,6,7-Tetrafluoro-2-(6,7,8,9-tetrahydro-5H-ben-
zocyclohepten-5-yl)-isoindole-1,3-dione (27). Compound
27 was from tetrafluorophthalic anhydride (400 mg/1.8
mmol) and amine 29 (360mg/1.8 mmol): mp 19–0
192 ꢂC; 1H NMR (DMSO-d6) d 1.2–1.3 (m, 1H), 1.7–1.8
(m, 1H), 1.9–2.0(m, 2H), 2.1–2.2 (m, 2H), 2.8–2.9 (m,
2H), 5.3–5.4 (m, 1H, NCH), 6.9–7.2 (m, 4H, ArH); 13C
NMR (DMSO-d6) d 162.7, 145.9, 144.4, 144.3, 142.4,
142.2, 140.9, 140.7, 140.4, 139.0, 130.0, 127.2, 126.1,
123.9, 54.8, 35.3, 32.6, 30.0, 26.9; EI m/z (rel intensity)
363.0(M +, 15), 144.0 (100). Anal. (C19H13F4NO2) C,
H, N; C: calcd, 62.81; found, 62.60; H: calcd, 3.61;
found, 3.51; N: calcd, 3.86; found, 3.84.
5.3.20.
4,5,6,7-Tetrafluoro-2-indan-1-yl-isoindole-1,3-
dione (23). Compound 23 was prepared from tetra-
fluorophthalic anhydride (400 mg/1.8 mmol) and 1-
aminoindan (240mg/1.8 mmol/.023 mL): mp 166–
168 ꢂC; 1H NMR (DMSO-d6) d 2.3–2.5 (m, 2H,
NCHCH2CH2), 2.9–3.0(m, 1H, CHNCH 2CH2), 3.1–
3.2 (m, 1H, NCHCH2CH2), 5.7 (t, J=7.45, 8.05 Hz,
1H, NCH), 7.1–7.3 (m, 4H, ArH); 13C NMR (DMSO-
d6) d 162.3, 143.5, 139.9, 128.0, 126.5, 124.7, 123.6, 54.8,
30.5, 29.0; EI m/z (rel intensity) 335.0(M +, 10), 116.0
(100). Anal. (C17H9F4NO2) C, H, N; C: calcd, 60.90;
found, 60.78; H: calcd, 2.71; found, 2.68; N: calcd, 4.18;
found, 4.16.
5.3.25. 6,7,8,9-Tetrahydrobenzocyclohepten-5-one oxime
(28). Amberlyst A-21 ion-exchange resin (1.25 g) was
added in one portion to a stirring solution of 1-benzo-
suberone (1.0g/6.24 mmol/.093 mL) and hydroxyl-
amine hydrochloride (870mg/12.48 mmol) in 20mL of
ethanol. The mixture was stirred at room temperature
for 24 h. The mixture was then filtered to remove the
resin and concentrated to give a yellow/white solid.
Ether (60mL) was added and the solution was filtered
to remove the excess hydroxylamine hydrochloride. The
ether solution was then dried (Na2SO4) and evaporated
to afford 28 as a light brown solid (82%). No further
purification was necessary: mp 97–99 ꢂC (lit.41 108 ꢂC);
1H NMR (CDCl3) d 1.6–1.8 (m, 4H), 2.7–2.8 (m, 4H),
7.1–7.5 (m, 4H, ArH), 9.3 (bs, 1H, NOH); 13C NMR
(CDCl3) d 162.4, 139.1, 135.6, 129.0, 128.5, 127.0, 126.2,
31.5, 25.6, 25.5, 21.1; APCI m/z (rel intensity) 175.8
(M+, 100).
5.3.21. 2-(1,2,3,4-Tetrahydronaphthalen-1-yl)-isoindole-
1,3-dione (24). Compound 24 was prepared from
phthalic anhydride (3.02 g/0.0204 mol) and 1,2,3,4-
tetrahydro-1-naphthylamine (3.0 g/0.0204 mol): mp
129–132 ꢂC (lit.40 127–128 ꢂC); H NMR (DMSO-d6) d
1
1.7–1.9 (m, 1H, CH), 2.0–2.2 (m, 2H, CH2), 2.3–2.5 (m,
1H, CH), 2.8–3.1 (m, 2H, CH2), 5.3 (dd, J=5.39 Hz,
1H, CH), 6.9–7.2 (m, 4H, ArH), 7.7–7.9 (m, 4H, ArH);
13C NMR (DMSO-d6) d 168.6, 138.2, 135.8, 135.5, 132.4,
129.9, 127.6, 127.0, 126.4, 124.1, 49.6, 29.8, 28.3, 22.8;
APCI m/z (rel intensity) 278.0(M +, 55), 131.2 (100).
5.3.26.
6,7,8,9-Tetrahydro-5H-benzocyclohepten-5-yl-
amine-hydrochloride (29). Compound 28 (900 mg/5.14
mmol) was dissolved in 20mL of absolute ethanol, and
10% Pd/C (50 mol%) was added. The mixture was
stirred under an atmosphere of H2 (g) for 3 days. The
used Pd/C was removed by filtration through Celite and
the solvent was removed in vacuo. The resulting clear vis-
cous oil/solid suspension was taken up in ether, and 6 N
HCl was added to precipitate the amine as the hydrochlo-
ride salt (100%). No further purification was necessary:
5.3.22. 4,5,6,7-Tetrafluoro-2-(1,2,3,4-tetrahydronaphtha-
len-1-yl)-isoindole-1,3-dione (25). Compound 25 was
prepared from tetrafluorophthalic anhydride (400 mg/
1.8 mmol) and 1,2,3,4-tetrahydro-1-naphthylamine (260
mg/1.8 mmol/0.26 mL): mp 160–162 ꢂC; 1H NMR
(DMSO-d6) d 1.7–2.2 (m, 4H, NCHCH2CH2CH2), 2.9–
3.1 (m, 2H, NCHCH2CH2CH2), 5.3–5.4 (q, J=5.44,
4.64, 5.84 Hz, 1H, NCH), 7.0–7.2 (m, 4H, ArH); 13C
NMR (DMSO-d6) d 162.4, 137.4, 134.0, 129.1, 126.9,
mp 268–271 ꢂC (lit.42 265–268 ꢂC); H NMR (DMSO-
1

334
S. M. Capitosti et al. / Bioorg. Med. Chem. 12 (2004) 327–336
d6) 0.8–0.9 (m, 1H), 1.1–1.5 (m, 4H), 1.6–1.7 (m, 1H),
2.2–2.4 (m, 2H), 3.9 (m, 1H, NCH), 6.6–6.8 (m, 4H, ArH),
8.3 (bs, 3H, NH3); 13C NMR (CDCl3) d 141.1, 137.8,
130.0, 128.0, 126.4, 124.4, 54.1, 35.2, 32.8, 28.3, 26.8;
APCI m/z (rel intensity) 144.9 (100), 197.9 (M+, 20).
7.2. IC50 Determination
A logarithmic concentration curve was made using
concentrations decreasing in half log steps between 300
and 0.1 mM, with each concentration tested in triplicate
both in the absence and in the presence of VEGF (as
described above). Negative controls were tested twice in
triplicate. Average absorbance values were divided by
average negative control values after subtracting aver-
age blank values from both and plotted versus concen-
tration. A linear regression was obtained and an IC50
value was calculated from that regression. For each
concentration, high and low absorbance values were
calculated by adding and subtracting, respectively, the
standard deviation. The resultant values were plotted
versus concentration to obtain high and low linear
regressions of which high and low values of the IC50
were determined. The absolute values of the deviations
from the original IC50 were averaged to determine the
standard deviation of the IC50. The percent error repre-
sents the standard error of the mean (SEM).
6. Biological methods
6.1. Endothelial cell culture
HMEC cells were grown in Gibco MCDB-131 media
supplemented with 2 mM l-glutamine (Gibco), 10%
fetal bovine serum (Gibco). The culture medium also
contained 50,000 IU/L penicillin (Gibco), 50 mg/L
streptomycin (Gibco), and 20 mg/L epidermal growth
factor (Sigma). Cells were routinely grown in filter-cap-
ped Nunc T25 and T75 flasks and kept at 37ꢁ0.5 ꢂC in
a humidified (90ꢁ1%) incubator under an air/CO2
(95%/5%) atmosphere.
Cells were re-fed 3 times per week and passaged once
per week. Upon passaging, cells received a 30ꢁ3 second
rinse with phosphate buffered saline pH 7.4, and were
then treated with Trypsin EDTA (0.05% Trypsin, 0.53
mM EDTA) (Gibco) for 2–7 min. The receiving cell
surface was previously treated at room temperature
with heated (37ꢁ0.5 ꢂC) 2% bovine gelatin serum
(Sigma) for 20–30 min, and then allowed to stand at
room temperature for ꢃ30min. Cell cultures were re-
fed every 24ꢁ3 h prior to seeding.
7.3. Prostate cancer cell culture
Du145, PC3 and LNCaP cells were grown in Gibco T-
Media (custom formulation) supplemented 5% fetal
bovine serum (Gibco) and 50,000 IU/L penicillin, 50
mg/L streptomycin (Gibco). Cells were routinely grown
in filter-capped Nunc T25 and T75 flasks and kept at
37ꢁ0.5 ꢂC in a humidified (90ꢁ1%) incubator under
an air/CO2 (95%/5%) atmosphere.
Cells were re-fed three times per week and passaged
once per week. Upon passaging, cells received a 30ꢁ3
second rinse with phosphate buffered saline pH 7.4, and
were then treated with 0.025% Trypsin without EDTA
(Gibco) for 2–7 min. Cell cultures were re-fed every
24ꢁ3 h prior to seeding.
7. Assay proliferation
Cells were seeded into a previously gelatin-treated 96-
well Nunc culture plate at approximately 5ꢄ103 cells
per well in MCDB-131 media as described above in
either the presence or absence of 10ng/mL vascular
endothelial growth factor (BD Biosciences). Drug dos-
ing occurred at 24ꢁ3 h, with redosing at 72ꢁ3 and
120ꢁ3 h. Drugs were dissolved in DMF at 375 mM and
diluted to yield final concentrations in half log steps
between 300 and 0.1 mM with a constant DMF concen-
tration of 0.2%.
7.4. Assay proliferation
Cells were seeded into a 96-well Nunc culture plate in T-
Media as described above at approximately 1ꢄ103 cells
per well for Du145 and PC3 lines, and approximately
2ꢄ103 cells per well for LNCaP’s. Drug dosing occurred
at 24ꢁ3 h, with redosing at 72ꢁ3 and 120ꢁ3 h. Drugs
were dissolved in DMF at 375 mM and diluted to yield
final concentrations in half log steps between 300 and
0.1 mM with a constant DMF concentration of 0.2%.
7.1. Crystal violet staining
Cells were fixed at 144ꢁ3 h with 1% gluteraldehyde in
PBS (Gibco) for 15 (ꢁ30s) min and then stained with 5
mg/L crystal violet (Fisher Scientific) in sterilized pur-
ified water for 15 min (ꢁ30s). No less than three rinses
(3–5) were performed to remove excess concentrated
stain with PBS (as mentioned above). The resulting
dilute stain was then removed through passive diffusion
in tap water immersion. At ꢃ3 h under minimal light
conditions cells received 250 mL/well Sorenson’s solu-
tion: 8.967 g trisodium citrate (Fisher Scientific), 19.5
mL 1 N HCl, 480 mL distilled water, in 500 mL 90%
Ethanol. Cells were shaken for 15 min at low speed and
read using a Thermo Labsystems Multiskan Ascent
plate reader at 492 nm immediately after completion on
the plate shaker (<1 min).
7.5. PSA ELISA
Using PSA ELISA kits from Yes Biotech, 50 mL media
samples from each test well (including all controls and
blanks) of the androgen dependent cell line, LNCaP,
were added to the antibody pre-coated Microtiter Plate
as well as standard samples all in duplicate at 144ꢁ3 h,
to which 50 mL of kit sample diluent was then added.
Plates were covered and shaken at moderate speed for 2
minꢁ20s and then incubated as described above for 30
minꢁ1 min. The ELISA plate was then rinsed with
approximately 300 mL of distilled water per well five
times. After rinsing, 100 mL of kit conjugate was added

S. M. Capitosti et al. / Bioorg. Med. Chem. 12 (2004) 327–336
335
to each well, the plate was covered, and incubated as
described immediately above. Rinsing was repeated
as described immediately above and then 50 mL of kit
substrate solution A was added to each well followed by
50 mL of kit substrate solution B. Plates were covered,
shaken at moderate speed for 30ꢁ3 s and incubated as
described immediately above for 15 minꢁ1 min. 100 mL
of kit stop solution was then added to each well and the
plate was covered and shaken at moderate speed for 5
minꢁ30s. Plates were immediately ( <1 min) read using
a Thermo Labsystems Multiskan Ascent plate reader at
450nm.
values were divided by average negative control values
after subtracting average blank values from both and
plotted versus concentration. A linear regression was
obtained and an IC50 value was calculated from that
regression. For each concentration, high and low absor-
bance values were calculated by adding and subtracting,
respectively, the standard deviation. The resultant
values were plotted versus concentration to obtain high
and low linear regressions of which high and low values
of the IC50 were determined. The absolute values of the
deviations from the original IC50 were averaged to
determine the average SEM.
7.6. Crystal violet staining
Acknowledgements
Cells were fixed at 144ꢁ3 h with 1% gluteraldehyde in
PBS (Gibco) for 15 min (ꢁ30s) and then stained with 5
mg/L crystal violet (Fisher Scientific) in sterilized puri-
fied water for 15 min (ꢁ30s). No less than three rinses
(3–5) were performed to remove excess concentrated
stain with PBS (as mentioned above). The resulting
dilute stain was then removed through passive diffusion
in tap water immersion. Cells received 250 mL/well Sor-
enson’s solution: 8.967 g trisodium citrate (Fisher Sci-
entific), 19.5 mL 1 N HCl, and 480mL distilled water in
500 mL 90% Ethanol, at ꢃ3 h under minimal light
conditions. Cells were then shaken for 15 min at low
speed and read using a Thermo Labsystems Multiskan
Ascent plate reader at 492 nm immediately after com-
pletion on the plate shaker (<1 min).
The American Chemical Society Division of Medicinal
Chemistry and Wyeth Pharmaceuticals are gratefully
acknowledged for support of a predoctoral fellowship
to S.M.C. We thank the Carilion Biomedical Institute,
University of Virginia Department of Chemistry, and
the Burger Center for Biological Sciences for financial
support. We thank Dr. Amy Tucker for the generous
donation of HMEC’s and Dr. Robert Sikes for the PC-3
and DU-145 cells. We also thank the National Cancer
Institute’s Developmental Therapeutics Program for
prostate cancer cell GI50’s.
References and notes
1. Li, J.; Zhang, Y.-P.; Kirsner, R. Micros. Res. Tech. 2003,
60, 107.
7.7. PSA determination
2. Reynolds, L. P.; Grazul-Bilska, A. T.; Redmer, D. A. Int.
J. Exp. Pathol. 2002, 83, 151.
3. Matter, A. Drug Discovery Today 2001, 6, 1005.
4. Bodolay, E.; Koch, A. E.; Kim, J.; Szegedi, Z. J. Cell. &
Mol. Med. 2002, 6, 357.
5. Noma, H.; Funatsu, H.; Yamashita, H.; Kitano, S.; Mis-
hima, H. K.; Hori, S. Arch. Ophthalmol. 2002, 120, 1075.
6. Fantl, W. J.; Rosenberg, S. Anti-angiogenesis as a Ther-
apeutic Strategy for Cancer. In Annual Reports in Medic-
inal Chemistry; Doherty, A. M., Ed.; Academic: San
Diego, CA, 2000; p 123.
7. Eriksson, T.; Bjorkman, S.; Hoglund, P. Eur. J. Clin.
Pharmacol. 2001, 57, 365.
8. Arora, M.; Wagner, J. E.; Davies, S. M.; Blazar, B. R.;
Defor, T.; Enright, H.; Miller, W. J.; Weisdorf, D. F. Biol.
Blood Marrow Transplant. 2001, 7, 265.
9. Gaziev, D.; Lucarelli, G. Exp. Opin. Invest. Drugs 2001,
10, 909.
10. Teo, S. K.; Resztak, K. E.; Scheffler, M. A.; Kook, K. A.;
Zeldis, J. B.; Stirling, D. I.; Thomas, S. D. Microbes
Infection 2002, 4, 1193.
A standard curve was plotted of ng PSA versus. O.D.450
nm using the 8 provided standards (subtracting the blank
from all subsequent values) and a linear trendline was
fitted to the values with b=0and acceptable R2 values
ꢃ99.9%. The O.D. values obtained for all ELISA test
wells were then compared to the trendline equation after
subtraction of the ELISA standard blank. The average
mass value of the cell blanks was then subtracted from
the averages of each set of test wells (twice in triplicate)
with acceptable standard deviations between duplicates
ꢅ0.1% of the actual value. The resultant normalized
masses were then divided by the negative control aver-
age normalized mass to give a mass to mass PSA per-
cent of control. That value was divided by the corollary
crystal violet ratio (described below): (average test
O.D.492 nm ꢀblank)/(average negative control O.D.492
ꢀblank). To calculate final values representing PSA
nm
secreted per cell, we plotted these ratios as percent of
control. This normalized the values such that all values
equal to one had no observable net effect on PSA
secretion, and all values deviating from one had a
respective correlation with increase or decrease in PSA
secretion.
11. Sheskin, J. Clin. Pharmacol. Ther. 1965, 6, 303.
12. Grosshans, E.; Illy, G. Int. J. Dermatol. 1984, 23, 598.
13. Kruse, F. E.; Joussen, A. M.; Rohrschneider, K.; Becker,
M. D.; Volcker, H. E. Graefes Archive for Clini. Exp.
Ophthalmol 1998, 236, 6 461.
14. D’Amato, R. J.; Loughnan, M. S.; Flynn, E.; Folkman, J.
Natil. Acad. Sci. U.S.A. 1994, 91, 4082.
7.8. IC50 Determination
15. Kenyon, B. M.; Browne, F.; D’Amato, R. J. Exp. Eye
Res. 1997, 64, 971.
16. (a) Chu, K. C.; Tarone, R. E.; Freeman, H. P. Cancer
2003, 97, 1507. (b) Harris, R.; Lohr, K. N. Ann. Int. Med.
2002, 137, 917.
A logarithmic concentration curve was made using
concentrations decreasing in half log steps between 300
and 0.1 mM, with controls and each concentration tested
in triplicate (as described above). Average absorbance

336
S. M. Capitosti et al. / Bioorg. Med. Chem. 12 (2004) 327–336
17. van Moorselaar, R. J. A.; Voest, E. E. Mol. Cell. Endo-
crinol. 2002, 197, 239.
28. Ballini, R.; Barboni, L.; Filippone, P. Chemistry Letters
1997, 5, 475.
18. Frydenberg, M.; Stricker, P. D.; Kaye, K. W. Lancet
1997, 349, 1681.
29. Bauer, K. S.; Dixon, S. C.; Figg, W. D. Biochem. Phar-
macol. 1998, 55, 1827.
19. Figg, W. D.; Dahut, W.; Duray, P.; Hamilton, M.;
Tompkins, A.; Steinberg, S. M.; Jones, E.; Premkumar,
A.; Linehan, W. M.; Floeter, M. K.; Chen, C. C.; Dixon,
S.; Kohler, D. R.; Kruger, E. A.; Gubish, E.; Pluda, J. M.;
Reed, E. Clin. Cancer Res. 2001, 7, 1888.
20. Ng, S. S. W.; Gutschow, M.; Weiss, M.; Hauschildt, S.;
Teubert, U.; Hecker, T. K.; Luzzio, F. A.; Kruger, E. A.;
Eger, K.; Figg, W. D. Cancer 2003, 63, 3189.
21. Shah, J. H.; Swartz, G. M.; Papathanassiu, A. E.; Tres-
ton, A. M.; Fogler, W. E.; Madsen, J. W.; Green, S. J. J.
Med. Chem. 1999, 42, 3014.
30. Filler, R., Kobayashi, Y., Yagupolskii, L. M., Eds.
Organofluorine Compounds in Medicinal Chemistry and
Biomedical Applications; Elsevier: Amsterdam, 1993.
31. Gordon, G. B.; Spielberg, S. P.; Blake, D. A.; Balasu-
bramanian, V. Natl. Acad. Sci. U.S.A. 1981, 78, 2545.
32. Perry, C. J.; Parveen, Z. J. Chem. Soc. Perkin Trans. 2
2001, 512.
33. Michman, M.; Patai, S.; Wiesel, Y. J. Chem. Soc.: Perkin
Trans. 1 1977, 1705.
34. Kjeldgaard, K. Arzneimittel-Forschung 1962, 12, 1207.
˜
35. Lopez-Alvarado, P.; Avendano, C.; Menendez, J. C. J.
´ ´
22. Miyachi, H.; Azuma, A.; Ogasawara, A.; Uchimura, E.;
Watanabe, N.; Kobayashi, Y.; Kato, F.; Kato, M.;
Hashimoto, Y. J. Med. Chem. 1997, 40, 2858.
23. Miyachi, H.; Azuma, A.; Kitamoto, T.; Hayashi, K.;
Kato, S.; Koga, M.; Sato, B.; Hashimoto, Y. Bioorg.
Med. Chem. Lett. 1997, 7, 1483.
Org. Chem. 1996, 61, 5865.
36. Rosenthal, A.; Weir, M. R. S. J. Org. Chem. 1963, 28, 3025.
37. Reynolds, R. D.; Anderson, G. L. J. Org. Chem. 1963, 28,
3223.
38. Bacon, R. G. R.; Karim, A. J. Chem. Soc. Perkin Trans. 1
1973, 272.
24. Niwayama, S.; Turk, B. E.; Liu, J. O. Journal of Medicinal
Chemistry 1996, 39, 3044.
39. Shibata, Y.; Sasaki, K.; Hashimoto, Y.; Iwasaki, S. Chem.
Pharm. Bull. 1996, 44, 156.
25. Niwayama, S.; Loh, C.; Turk, B. E.; Liu, J. O.; Miyachi,
H.; Hashimoto, Y. Bioorg. and Med. Chem. Lett. 1998, 8,
1071.
26. Shibata, Y.; Sasaki, K.; Hashimoto, Y. Biochem. Biophys.
Res. Comm. 1994, 205, 1992.
27. Hartmann, K.-H.; Troll, T. Tetrahedron 1995, 51, 4655.
40. Ghislandi, V.; La Manna, A. Farmaco Ed. Sci. 1976, 31,
489.
41. Khanna, J. M.; Chak, I. M.; Anand, N. Indian J. Chem.
1967, 5, 347.
42. Vejdelek, Z. J.; Protiva, M. Coll. Czech. Chem. Commun.
1971, 36, 1611.