Therapeutic potential of sulindac hydroxamic acid against human pancreatic and colonic cancer cells

Article abstract of DOI:10.1016/j.ejmech.2010.08.019

The non-steroidal anti-inflammatory drug (NSAID) sulindac exhibits cyclooxygenase (COX)-dependent and COX-independent chemopreventive properties in human cancer. The present study was aimed at investigating whether the hydroxamic acid substitution for the carboxylic acid group could enhance the in vitro antitumor and antiangiogenic activities of sulindac. Characterization tools used on this study included analyses of cell viability, caspase 3/7 induction, DNA fragmentation, and gene expression. Our findings demonstrate that the newly synthesized hydroxamic acid derivative of sulindac and its sulfone and sulfide metabolites were characterized by a good anticancer activity on human pancreatic and colon cancer cells, both in terms of potency (IC ₅₀ mean values from 6 ± 1.1 μM to 64 ± 1.1 μM) and efficacy (E_max of ～100%). Hydroxamic acid derivatives trigger a higher degree of apoptosis than carboxylic acid counterparts, increase bax/bcl-2 expression ratio and induce caspase 3/7 activation. Most notably, these compounds significantly inhibit proangiogenic growth factor-stimulated proliferation of vascular endothelial cell (HUVEC) at sub-micromolar concentrations. Our data also provide evidence that the COX-active metabolite of sulindac hydroxamic acid were the most active of the series and selective inhibition of COX-1 but not COX-2 can mimic its effects, suggesting that COX inhibition could only play a partial role in the mechanism of compound action. In conclusion, these data demonstrate that substitution of the carboxylic acid group with the hydroxamic acid moiety enhances in vitro antiproliferative, proapoptotic and antiangiogenic properties of sulindac, therefore increasing the therapeutic potential of this drug.

Full text of DOI:10.1016/j.ejmech.2010.08.019

European Journal of Medicinal Chemistry 45 (2010) 5100e5107
Contents lists available at ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Therapeutic potential of sulindac hydroxamic acid against human pancreatic and
colonic cancer cells
,
b,
Stefano Fogli a , Irene Banti 1, Fabio Stefanelli a, Luca Picchianti a, Maria Digiacomo b, Marco Macchia b,
*
Maria Cristina Breschi a, Annalina Lapucci b
a Department of Psychiatry, Neurobiology, Pharmacology and Biotechnologies, University of Pisa, Via Bonanno 6, 56126 Pisa, Italy
b Department of Pharmaceutical Sciences, University of Pisa, Via Bonanno 6, 56126 Pisa, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
The non-steroidal anti-inflammatory drug (NSAID) sulindac exhibits cyclooxygenase (COX)-dependent
and COX-independent chemopreventive properties in human cancer. The present study was aimed at
investigating whether the hydroxamic acid substitution for the carboxylic acid group could enhance the
in vitro antitumor and antiangiogenic activities of sulindac. Characterization tools used on this study
included analyses of cell viability, caspase 3/7 induction, DNA fragmentation, and gene expression. Our
findings demonstrate that the newly synthesized hydroxamic acid derivative of sulindac and its sulfone
and sulfide metabolites were characterized by a good anticancer activity on human pancreatic and colon
Received 9 June 2010
Received in revised form
29 July 2010
Accepted 9 August 2010
Available online 12 August 2010
Keywords:
Sulindac
cancer cells, both in terms of potency (IC50 mean values from 6 ꢀ 1.1
m
M to 64 ꢀ 1.1
mM) and efficacy
(Emax of w100%). Hydroxamic acid derivatives trigger a higher degree of apoptosis than carboxylic acid
counterparts, increase bax/bcl-2 expression ratio and induce caspase 3/7 activation. Most notably, these
compounds significantly inhibit proangiogenic growth factor-stimulated proliferation of vascular
endothelial cell (HUVEC) at sub-micromolar concentrations. Our data also provide evidence that the
COX-active metabolite of sulindac hydroxamic acid were the most active of the series and selective
inhibition of COX-1 but not COX-2 can mimic its effects, suggesting that COX inhibition could only play
a partial role in the mechanism of compound action. In conclusion, these data demonstrate that
substitution of the carboxylic acid group with the hydroxamic acid moiety enhances in vitro anti-
proliferative, proapoptotic and antiangiogenic properties of sulindac, therefore increasing the thera-
peutic potential of this drug.
Hydroxamic acid
Cytotoxicity
Apoptosis
Angiogenesis
Ó 2010 Elsevier Masson SAS. All rights reserved.
1. Introduction
should be able to interfere with different elements which play
a pivotal role into the tumorehost interface circuit.
Inflammatory processes have been shown to enhance carcino-
genesis and tumor progression in multiple ways including
production of reactive oxygen species and induction of vascular
endothelial growth factor. Following the initiation of solid tumors,
the cross-talk between transformed epithelial and stromal cells has
a key role in promoting cancer progression. For example, in stromal
cells, proinflammatory cytokines induce up-regulation of COX-2
which, in turn, may contribute to cancer development by producing
adhesion molecules and activating neoangiogenesis, particularly
through the production of prostanoids [1]. These notions highlight
that effective cancer chemopreventive and/or therapeutic agents
Sulindac is a non-steroidal anti-inflammatory drug (NSAID) of
the arylalkanoic acid class that, in vivo, is irreversibly oxidized to
sulindac sulfone or reversibly converted into its anti-inflammatory
active compound, sulindac sulfide. Sulindac and sulindac sulfone
are devoid of any activity as COX inhibitors, while most of the
biological effects of the sulfide metabolite are dependent upon
inhibition of both cyclooxygenase-1 (COX-1) and COX-2 activities
and reduction in prostaglandin (PG) synthesis [2]. Compelling
evidence has been provided on the therapeutic potential of sulin-
dac as chemopreventive agent in colorectal cancer and both COX-
dependent and COX-independent mechanisms appear to
contribute to such an effect [3]. In particular, it has been recently
demonstrated that induction of cytochrome P450 and carcinogen
detoxification enzymes such as NAD(P)H:quinine oxidoreductase
and glutathione S-transferase may account for the chemo-
preventive action of sulindac and its metabolites in human colon
* Corresponding author. Tel.: þ39 050 2219539; fax: þ39 050 2219609.
E-mail address: s.fogli@farm.unipi.it (S. Fogli).
1
This author contributed equally to this work.
0223-5234/$ e see front matter Ó 2010 Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.ejmech.2010.08.019

S. Fogli et al. / European Journal of Medicinal Chemistry 45 (2010) 5100e5107
5101
cancer cells [4]. The antitumor properties of sulindac were mostly
due to the ability of sulfone and sulfide metabolites to negatively
interfere with Ras signaling in tumor cells [5e7] and with angio-
genetic pathways in the surrounding tissue [8]. Finally, several lines
of evidence also indicate that sulindac is able to potentiate the
effect of different chemotherapeutic agents including cisplatin,
cyclophosphamide, melphalan, carmustine and doxorubicin [9].
Studies have shown that most, if not all, NSAIDs cause increased
risk of heart attack in humans and such an effect can be related to
the inhibition of the inducible COX-2 isoform in the cardiac tissue
[10]. Noteworthy, sulindac confers ischemic protection, in vitro and
in vivo, through late preconditioning mechanisms which did not
involve its role as a COX inhibitor [11].
It has been suggested that incorporating a chelating group
(hydroxamic acid) into the structure of conventional COX inhibitors
resulted in the acquisition of antioxidant properties which might
increase the therapeutic benefit of these drugs because of impli-
cation of free radical production in the pathogenesis of cancer and
ischemic-reperfusion injury [12]. Furthermore, recent findings
demonstrate that sulindac and some hydroxamic acids synergisti-
cally induce apoptosis in human cancer cells [13], suggesting that
the chemical modification of sulindac structure might be a reason-
able strategy for the development of new compounds with
enhanced antitumor activity. Overall, these data prompted us to
synthesize and characterized the in vitro antitumor, proapoptotic
and antiangiogenic properties of sulindac hydroxamic acid and its
sulfone and sulfide metabolites, in order to take into account all
chemical species potentially involved in the mechanism of drug
action, in vivo. Sulindac, sulindac sulfone and sulindac sulfide were
evaluated for comparison (Fig. 1).
Scheme 1. Synthesis of sulfone and sulfide metabolites of sulindac. Reagents and
conditions: (i) oxone, THF/MeOH 1/1 v/v, 0e25 ꢁC, 24 h; (ii) Ph3P, TiCl4, THF, rt, 24 h.
2.2. Pharmacology
2.2.1. Cytotoxicity of compounds against human cancer cells
The antitumour effect in vitro of test compounds was assessed
against human cancer cell lines derived from colon (COLO320) and
pancreas (MIA PaCa-2). Overall, COLO320 cells were most respon-
sive and exposure of the cells to compounds caused a concentra-
tion-dependent decrease in cell viability with mean 50% inhibition
(IC50) in the micromolar range of concentrations (Table 1). Within
carboxylic acid compounds (aec), only sulindac sulfide (c) exhibi-
ted some anticancer activity, particularly against COLO320 cells
(Fig. 2A and C; Table 1). The hydroxamic acid derivatives (1e3) were
more active than their carboxylic acid counterparts, in terms of
potency (Table 1), and proven to have a 100% efficacy in both cell
lines tested (Fig. 2B and D).
2. Results
2.1. Chemistry
The acids b,c were prepared starting from sulindac (a) [14] by
using the following steps (Scheme 1) according to the literature for
the same compounds [15]. The reaction of a (commercially avail-
able), with the oxone, in a mixture of THF/MeOH 1:1 at room
temperature afforded sulindac sulfone (b). The reaction of a in the
presence of Ph3P and TiCl4 (solution 1 M in CH2Cl2) in THF at room
temperature afforded sulindac sulfide (c).
2.2.2. Mechanisms of cell death induced by compounds
Since cytotoxicity assay did not indicate a specific cellular death
mechanism, we investigated the ability of selected compounds to
induce internucleosomal degradation of genomic DNA (a hallmark
The hydroxamic acids 1e3 were prepared starting from a [14] by
using the following steps (Scheme 2) according to the literature for
compounds with similar structure [15]. The reaction of a with
TBDMSiONH2 (O-(tert-butyl-dimethylslilyl)hydroxilamine), in the
presence of EDCI (N-3,dimethylaminopropyl-N-ethylcarbodimmide
hydrochloride), gave the O-silylated hydroxamates 4e6, which
were transformed by acid cleavage to the corresponding hydroxamic
acids 1e3 [16].
Scheme 2. Synthesis of hydroxamic acid derivatives of sulindac. Reagents and
conditions: (iii) TBDMSONH2, EDCI, dry CH2Cl2, 0e25 ꢁC, 24 h; (iv) TFA, dry CH2Cl2,
0
Fig. 1. Chemical structures of sulindac (a), sulfone (b) and sulfide (c) metabolites and
their hydroxamic acid derivatives (1e3).
ꢁC, 5 h.

5102
S. Fogli et al. / European Journal of Medicinal Chemistry 45 (2010) 5100e5107
Table 1
proteins involved in apoptosis regulation, Bax (proapoptotic) and
IC50s
(m
M, mean ꢀ SEM; n ¼ 3) of test compounds on human cancer cell lines.
Bcl-2 (antiapoptotic), after treatment of cancer cells with c and its
hydroxamic acid derivative, 3 (i.e., the most active compounds
tested). Although compounds c and 3 induced different modulation
of bax and bcl-2 gene expression in MIA PACa-2 cells, the bax:bcl-2
(proapoptotic:antiapoptotic) expression ratio after treatments was
similar to each other and to that of controls (Fig. 4A). Noteworthy,
such an effect was in line with the level of caspase 3/7 induction
observed in MIA PaCa-2 cells after treatment with these two
compounds. In COLO320 cells, bax:bcl-2 expression ratio was
almost superimposable in c and control samples, while it increased
by 130% after treatment with 3 (Fig. 4B). These findings fit well with
the present data showing the more pronounced ability of 3 than c
to promote caspase 3/7 induction in COLO320 cell line.
Compounds
Cell line
MIA PaCa-2
COLO320
a
b
c
1
2
3
>200
>200
146 ꢀ 3.6
64.1 ꢀ 1.1
44.4 ꢀ 1.1
32 ꢀ 1.3
>200
>200
38.7 ꢀ 1.2a
62.5 ꢀ 1.2
26.2 ꢀ 1.2a
6.0 ꢀ 1.1a
a
Changes in the mean IC50 values obtained for each cell line that are significant
(p < 0.001; Student’s t-test).
of apoptosis). While the amount of apoptotic DNA generated upon
exposure of both cancer cell lines to sulindac and its metabolites at
2.2.3. Involvement of COX enzymes in the effect induced by
hydroxamic acid derivatives
200 mM (i.e., the maximum concentration tested in cell viability
assay) did not differ significantly from that of untreated controls,
hydroxamic acid derivatives were able to induce apoptosis (Fig. 3A
and B). In particular, DNA degradation and formation of oligonu-
cleosomal fragments in the cytoplasmic fraction of colonic cells was
greater than those observed in pancreatic cancer cells (up to 6-
versus 15-fold over control, respectively). Noteworthy, 2 caused cell
death mainly by apoptosis in both cell lines (apoptosis to necrosis
ratio of w4:1), whereas after treatment with 3, a switch in the cell
death mode from apoptosis to necrosis was observed in MIA PaCa-2
and COLO320 cells (ratio of w1:1 and 1:12, respectively) (Fig. 3A
and B) showing that the extent and the type of cell death were
dependent upon the cell line and compound tested.
The possible induction of caspase 3/7 (two enzymes involved in
the effector phase of apoptosis) by compounds was also investi-
gated. Treatment of MIA PACa-2 cells with hydroxamic acid deriv-
atives was associated with a significantly increase in the extent of
caspase 3/7 stimulation up to 5-fold over control. In COLO320 cells,
caspase activation was consistently observed only after incubation
with 3, while the effect of other compounds was negligible (Fig. 3C).
In the same experimental conditions, we evaluated possible
alterations in the expression of genes that code for two important
Expression of constitutive COX-1 and inducible COX-2 isoforms
was confirmed by RT-PCR. As expected, relatively low COX-1 mRNA
levels was detected in both cell lines, while COX-2 was expressed
only in the human colonic cells (Fig. 5A and B). Experiments with
selective COX inhibitors on human cancer cells were performed to
investigate the possible contribute of these enzymes to the mech-
anism of compound action. The concentrationeresponse curves
showed that SC560 (COX-1) was active with a potency similar to
that of 3, particularly in COLO320 cells, while no effect was
observed after treatment with DFU (COX-2) up to 200
mM (Fig. 5C
and D; Table 2). These data suggest that the inhibition COX-1 but
not COX-2 might play a partial role in the mechanism of compound
action.
To support such a hypothesis, we also considered the possibility
that cytotoxic effect induced by compounds can be due to the
increased levels of the prostaglandin precursor arachidonic acid
(AA) resulting from inhibition of COX, as previously reported [17].
We found that AA was cytotoxic against both pancreatic and colonic
cell lines (Fig. 6A). In particular, AA at 200
mM for 48 h caused cell
death with an apoptosis to necrosis ratio of 5:1 and 1:2 in MIA
Fig. 2. Concentration-response curves of test compounds in MIA PaCa-2 (A and B) and COLO320 (C and D) cells after 72 h. Data are expressed as mean ꢀ S.E.M (n ¼ 3).

S. Fogli et al. / European Journal of Medicinal Chemistry 45 (2010) 5100e5107
5103
Fig. 3. Apoptotic and necrotic cell death by test compounds 200 mM in MIA PaCa-2 (A) and COLO320 (B) cells after 48 h. Induction of caspase 3/7 activity by test compounds 200 mM
in MIA PaCa-2 and COLO320 cells after 24 h (C). Data are expressed as mean ꢀ S.E.M (n ¼ 3).
PaCA-2 and COLO320 cells, respectively (Fig. 6B), while caspase 3/7
activity was not affected by treatment (Fig. 6C).
promote an increase in the bax/bcl-2 expression ratio and induc-
tion of caspase 3/7, two enzymes involved in the effector phase of
apoptosis. Of note, such an effect was associated with inter-
nucleosomal DNA fragmentation suggesting that, at least in this
specific cancer cell type, caspase activation may contribute to the
proapoptotic action of compounds. These findings appear to be
relevant since bcl-2 overexpression confers resistance to sulindac-
induced apoptotic signaling in human colon cancer cells [18].
Growing evidence supports the hypothesis that the develop-
ment of cancer is angiogenesis-dependent. Beyond its role in the
metastatic spread and growth of advanced tumors, angiogenesis
could also represent a useful target for chemoprevention since the
acquisition of a blood supply is critical for the growth of an early
neoplastic lesion [19]. We demonstrated that sulindac hydroxamic
acid and metabolites exerted remarkable in vitro antiangiogenic
activity. Particularly, hydroxamic acid derivatives of sulindac
sulfone and sulfide were able to inhibit 50% of HUVEC growth
2.2.4. Inhibition of angiogenesis
Growth inhibition studies were performed on HUVEC grown in
the presence of the angiogenic factors, bFGF and VEGF. Under this
conditions, all test compounds but a, significantly inhibited cell
proliferation after 96-h exposure (P < 0.05) and such an effect
occurred at 0.1 mM (Fig. 7A). Sulindac hydroxamic acid (1) and its
metabolites (2 and 3) were more active than their carboxylic acids
counterparts and the order of efficacy was 3 > 2 > 1 (Fig. 7A). Cell
viability analyses performed by the trypan blue exclusion test
indicated that no cytotoxicity occurred in these experimental
conditions (data not shown). Finally, high expression levels of both
constitutive COX-1 and inducible COX-2 isoforms were demon-
strated in HUVEC by RT-PCR (Fig. 7B).
3. Discussion
stimulated by proangiogenic factors at 0.1
mM and such an effect
was higher than that observed for sulindac and metabolites.
With regard for the involvement of COX enzymes in the
molecular mechanism of action of sulindac hydroxamic acid, it is
noteworthy that attaching hydroxamic acid to COX inhibitors did
not affect their ability to arrest prostaglandin synthesis [12].
Therefore, we can assume that the hydroxamic acid derivative of
In the present study, we demonstrated that substitution of the
carboxylic acid group with the hydroxamic acid moiety substan-
tially enhances in vitro proapoptotic and antiangiogenic activities
of sulindac and sulindac metabolites. Concerning the molecular
mechanism of action, hydroxamic acid derivatives were able to
Fig. 4. Effect of c and its hydroxamic derivative, 3, at 200
mM for 24 h on bax and bcl-2 mRNA expression levels in MIA PaCa-2 (A) and COLO320 (B) cells. Levels of mRNA were
measured by semi-quantitative RT-PCR, and normalized to GAPDH mRNA. Data are expressed as mean ꢀ S.E.M (n ¼ 3).

5104
S. Fogli et al. / European Journal of Medicinal Chemistry 45 (2010) 5100e5107
experimental setting since selective COX-2 inhibition did not affect
cell proliferation and sulindac sulfide (the COX-active metabolite of
sulindac) was also active against MIA PaCa-2 cells. Pancreatic MIA
Paca-2 cells were less responsive than COLO320 cells to compound-
induced cytotoxicity, a phenotypic feature most probably due to the
intrinsic apoptotic resistance of this specific cell type, as previously
reported [21]. Therefore, we can speculate that the antiproliferative
and proapoptotic effects of the COX-active compounds tested in the
present study (sulindac sulfide and its hydroxamic acid derivative,
3) might be partially mediate by COX-1 inhibition, since the
selective COX-1 inhibitor SC560 can mimic their effect, both in
terms of potency and efficacy. Evidence in line with this notion was
reported on the effect of low-dose aspirin (COX-1 inhibitor) against
metachronous polyps and on the significant role of COX-1 poly-
morphisms in determining the chemopreventive potential of
NSAIDs [22].
It has been proposed that a potential COX-mediated mechanism
responsible for the tumor suppressive effects of sulindac might
result from an accumulation of the prostaglandin precursor
arachidonic acid, which stimulates the conversion of sphingo-
myelin to ceramide, a known mediator of apoptosis [17]. Ceramide,
in turn, mediated caspase-independent programmed cell death
[23]. Although we did not directly measure changes in intracellular
arachidonic acid concentration following treatment with 3, exog-
enous arachidonic acid induced cytotoxicity and caused apoptosis
in colon and pancreatic cancer cells, independently from caspase 3/
7 activation. As previously demonstrated for sulindac sulfide in
human colon cancer cells [17], our findings suggest that COX-1
inhibition and the consequent arachidonic acid increase might
partially account for 3 cytotoxicity.
With regard for the clinical relevance of our results, we
demonstrate that the cytotoxic activity against pancreatic and
colon cancer cells and the in vitro antiangiogenic effect induced by
sulindac hydroxamic acid and metabolites occur at concentrations
lower than carboxylic acid counterparts (i.e., in the low micromolar
range). Plasma levels of sulindac have been reported to be up to
Fig. 5. Expression levels of COX-1 and COX-2 mRNA in MIA PaCa-2 (A) and COLO320
(B) cells. Concentration-response curves of the selective COX inhibitors, SC560 (COX-1)
and DFU (COX-2) in MIA PaCa-2 (C) and COLO320 (D) cells after 72 h. Data are
expressed as mean ꢀ S.E.M (n ¼ 3).
sulindac sulfide, 3, is able to maintain the COX inhibitor activity.
With this in mind, our findings suggest that both COX-dependent
and COX-independent effects might contribute to the mechanism of
action of test compounds. This conclusion is based on the following
lines of evidence: (i) sulindac sulfide (the COX-active metabolite of
sulindac) and its hydroxamic acid derivative, 3, were the most
active compounds tested within carboxylic and hydroxamic acids
series, respectively; (ii) sulindac sulfide and 3 were active against
both COX-2-negative (MIA PACa-2) and COX-2-positive (COLO320)
cells; and (iii) selective inhibition of COX-1 but not COX-2 can mimic
their effects both in terms of potency and efficacy.
Several lines of evidence indicate that COX-2 isoform and its
products are generally highly expressed in premalignant and
malignant lesions of intestinal epithelia and they are involved in
stimulation of tumor growth and metastases [19]. Our findings
demonstrated that COLO320 cells were more sensitive to the
cytotoxic action of test compounds than MIA PaCa-2 cells. While
the COX-1 isoform was expressed at low levels in both cancer cell
lines, differences in COX-2 gene expression were observed.
Specifically, COLO320 cells were characterized by high levels of
COX-2 mRNA whereas, in line with literature data [20], COX-2 gene
was not expressed in MIA PaCa-2 cells. Although these data suggest
a possible relationship between COX-2 expression and in vitro drug
response, the role of the inducible isoform of COX in the mainte-
nance of the COLO320 phenotype appears to be negligible in our
50 mM after oral administration and the concentration of its sulfide
metabolite is higher in the intestines than in other organs, due to
the enterohepatic circulation and to the metabolic activation of
parent drug by gut flora [24]. Therefore it is conceivable that
sulindac hydroxamic acid and metabolites can reach a threshold
level required for biologically activity in the tumor microenviron-
ment, in vivo.
In conclusion, the chemical transformation of sulindac into
hydroxamic acid derivative enhances its proapoptotic and anti-
angiogenic properties, thus improving the therapeutic potential of
the drug against human cancer.
4. Experimental
4.1. Chemistry
4.1.1. Materials and physical measurements
Melting points were determined on a Kofler hot stage apparatus
and are uncorrected. Infrared (IR) spectra for comparison of
compounds were recorded on a Mattson 1000 FTIR spectrometer.
Nuclear magnetic resonance (1H NMR) spectra were recorded on
a Varian Gemini 200 (200 MHz) in a ca. 2% solution of CDCl3 for all
compounds. Peak positions are given in parts per million (ppm,
Table 2
IC50s
(m
M, mean ꢀ SEM; n ¼ 3) of selective COX inhibitors on human cancer cell lines.
Compounds
Cell line
d
units). The proton magnetic resonance assignments were estab-
MIA PaCa-2
COLO320
lished on the basis of the expected chemical shifts and the multi-
plicity of the signals. Reactions were routinely monitored by
thin-layer chromatography (TLC) on 0.25 mm silica gel plates
(Merck 60F254) and hydroxamic acids were visualized with FeCl3
aqueous solution. Flash chromatography was carried out through
SC560
DFU
102 ꢀ 1.2
>200
5.6 ꢀ 1.1a
>200
SC560: selective COX-1 inhibitor; DFU: selective COX-2 inhibitor.
a
Changes in the mean IC50 values obtained for each cell line that are significant
(p < 0.001; Student’s t-test).

S. Fogli et al. / European Journal of Medicinal Chemistry 45 (2010) 5100e5107
5105
Fig. 6. Cytotoxicity (A), apoptosis/necrosis (B) and caspase 3/7 induction (C) after treatment of human cancer cells with arachidonic acid (AA) 200
respectively. Data are expressed as mean ꢀ S.E.M (n ¼ 3).
mM for 72, 48 and 24 h,
glass columns containing 40e63
m
m silica gel (MachereyeNagel
Na2SO4. This organic layer was filtered and evaporated under
reduced pressure, and then the crude product was triturated with
Et2O to remove the excess of triphenylphosphine, filtered and
evaporated in vacuo to yield a crude product that was purified by
titration with Et2O to give the sulfone c as a yellow solid. Compound
Silica Gel 60). Solvents and reagents were obtained from
commercial sources in the appropriate grade and were used
without further purification unless otherwise indicated. Reactions
were run under an argon atmosphere. Elemental analyses were
carried out by our analytical laboratory and were consistent with
theoretical values to within ꢀ0.4%.
c: mp: 186e190 ꢁC; yield 75%; 1H NMR (CDCl3,
d ppm): 7.73e7.30
(m, 4H); 7.14 (s, 2H); 6.91e6.86 (m, 1H); 6.63e6.53 (m, 1H); 3.59 (s,
2H); 2.54 (s, 3H); 2.20 (s, 3H). Anal. Calcd for C20H17FO2S: C, 70.57;
H, 5.03. Found: C, 70.76; H, 5.24.
4.1.2. Procedures for the synthesis of sulindac metabolites
4.1.2.1. (Z)-2-(6-Fluoro-2-methyl-3-(4-(methylsulfonyl)benzylidene)-
3H-inden-1-yl)acetic acid (b). A solution of oxone (1.7 mmol) in H2O
(3 ml), was added portionwise to stirred and cooled (0 ꢁC) solution of
sulindac (a) (1.4 mmol) inTHF/MeOH 1/1 v/v (10 ml). After stirring at
room temperature for 24 h, the mixture was extracted with AcOEt,
washed with water (20 ml) and the organic phase was dried and
evaporated in vacuo to yield the corresponding sulfone derivative, b.
Compound b was yellow solid, mp: 199e200 ꢁC, yield 75%; 1H NMR
(CDCl3, d ppm): 8.03e7.98 (d, J ¼ 8.2 Hz, 2H); 7.72e7.67 (d, J ¼ 8.2 Hz,
2H); 7.14e7.06 (m, 2H);6.91e6.86 (m,1H); 6.62e6.52 (m,1H) 3.60 (s,
2H); 3.13 (s, 3H); 2.21 (s, 3H). Anal. Calcd for C19H17FO4S: C, 63.32, H,
4.75. Found: C, 63.17; H, 3.58.
4.1.3. General procedure for the synthesis of O-TBDMS acid
hydroxyamides 4e6
1-[3-(Dimethylamino)propyl]-3-ethyl carbodiimide hydrochlo-
ride (EDCI) was added portionwise (1 mmol) to stirred and cooled
(0 ꢁC) solution of the carboxylic acids aec (1 mmol) in dry CH2Cl2
(18 ml). After stirring at room temperature for 20 h, the mixture
was washed with water (20 ml) and the organic phase was dried
and evaporated in vacuo. The residue was purified by flash chro-
matography to yield pure silyl esters 4e6 as oils.
4.1.3.1. (Z)-2-(6-Fluoro-2-methyl-3-(4-(methylsulfinyl)benzylidene)-
3H-inden-1-yl) acetic acid (tert-butyldimethylsilyl)hydroxyamide
(4). The title compound was prepared from carboxylic acid
a following the general procedure. The crude product was purified
by flash chromatography on silica gel using hexane-ethyl acetate 4/
4.1.2.2. (Z)-2-(6-Fluoro-2-methyl-3-(4-(methylthio)benzylidene)-3H-
inden-1-yl)acetic acid (c). To a solution of a (0.20 mmol) and TiCl4
(0.3 mmol) in THF (2.5 ml) was added a solution of triphenylphos-
phine (0.24 mmol) in THF (1.5 ml) and the mixture was stirred at
appropriate temperature under an argon atmosphere. The reaction
was quenched with saturated sodium hydrogen carbonate (10 ml),
and the mixture was extracted with ether (3 ꢂ 10 ml). The combined
ether layers were washed with brine (1 ꢂ 10 ml) and dried over
6 v/v to give 4 as a yellow oil, yield 40%; 1H NMR (CDCl3,
d ppm):
7.71e7.68 (m, 4H); 7.18 (s, 2H); 6.87e6.83 (m, 1H); 6.63e6.54 (m,
1H); 3.51 (s, 2H); 2.81 (s, 3H); 2.21 (s, 3H); 0.91 (s, 9H); 0.14 (s, 6H).
Anal. Calcd for C25H32FNO3SSi: C, 63.40; H, 6.82; N, 2.96. Found: C,
63.57; H, 6.98; N, 3.14.
Fig. 7. Effect of test compounds on HUVEC proliferation after 96-h exposure at 0.1
(n ¼ 3). **P < 0.01 (a versus 1), ***P < 0.001 (b versus 2 and c versus 3).
mM (A) and expression levels of COX-1 and COX-2 mRNA (B). Data are expressed as mean ꢀ SEM

5106
S. Fogli et al. / European Journal of Medicinal Chemistry 45 (2010) 5100e5107
4.1.3.2. (E)-2-(6-Fluoro-2-methyl-3-(4-(methylsulfonyl)benzyli-
dene)-3H-inden-1-yl) acetic acid (tert-butyldimethylsilyl)hydrox-
yamide (5). The title compound was prepared from carboxylic acid
a following the general procedure. The crude product was purified
by flash chromatography on silica gel using hexane-ethyl acetate 4/
4.2.2. Cell viability assay
Pancreatic or colon cancer cells (2e5 ꢂ 103/well) were seeded
into 96-well microtiter plates and received compounds from 5 to
200
mM for 72 h. Following drug exposure, cell viability was
assessed using a method based on the cleavage of the tetrazolium
salt WST-1 to formazan by mitochondrial dehydrogenase activity
(Cell proliferation reagent WST-1; Roche, Mannheim, Germany).
Inhibition of proliferation was reported as the percentage reduction
of UV absorbance of treated cells versus control cultures and the
concentration of compounds that decreased cell viability by 50%
(IC50) was calculated by non-linear least squares curve fitting
(GraphPad Software, San Diego, CA, USA). DMSO concentration in
the culture medium never exceeded 0.2%.
6 v/v to give 5 as a yellow oil, yield 40%; 1H NMR (CDCl3,
d ppm):
7.97e7.93 (m, 2H); 7.72e7.68 (m, 2H); 7.19e7.10 (m, 2H); 6.94e6.83
(m, 1H); 3.58 (s, 2H); 3.11 (s, 3H); 2.21 (s, 3H); 0.91 (s, 9H); 0.14 (s,
6H). Anal. Calcd for C25H32FNO4SSi: C, 61.32; H, 6.59; N, 2.86.
Found: C, 61.13; H, 6.42; N, 2.70.
4.1.3.3. (Z)-2-(6-Fluoro-2-methyl-3-(4-(methylthio)benzylidene)-3H-
inden-1-yl) acetic acid (tert-butyldimethylsilyl)hydroxyamide (6). The
title compound was prepared from carboxylic acid a following the
general procedure. The crude product was purified by flash chroma-
tography on silica gel using hexane-ethyl acetate 4/6 v/v to give 6 as
4.2.3. Caspase activity assay
Enzyme activity was assayed by the Apo-ONEÔ Homogeneous
Caspase 3/7 assay (Promega, Madison, WI, USA). Briefly, cells were
a yellow oil, yield 40%; 1H NMR (CDCl3,
d ppm): 7.71e7.30 (m, 4H); 7.15
(s, 2H); 6.89e6.83 (m,1H); 6.63e6.60 (m,1H); 3.51 (s, 2H); 2.55 (s, 3H);
2.21 (s, 3H); 0.91 (s, 9H); 0.14 (s, 6H). Anal. Calcd for C26H32FNO2SSi: C,
66.49; H, 6.87; N, 2.98. Found: C, 66.68; H, 7.10; N, 3.16.
seeded at 5 ꢂ 103/well and treated with compounds at 100
mM for
24 h. Subsequently, the Caspase 3/7 assay substrate was added and
the fluorescence was measured by spectrofluorimeter at excitation
and emission wavelengths of 485 and 530 nm, respectively. Values
were expressed as ratio between fluorescent signals generated in
cells treated with compounds and those produced in untreated
cells (vehicle alone).
4.1.4. General procedureforthepreparation ofacid hydroxyamides 1e3
Trifluoroacetic acid (4.4 mL, 57 mmol) was added dropwise to a stir-
redand cooled solution(0 ꢁC)of4e6(1mmol)indryCH2Cl2.Thesolution
was stirred under these reaction conditions for 5 h and the solvent was
removed in vacuo to give the hydroxyacetamide derivatives (1e3).
4.2.4. DNA fragmentation assay
Cells were treated for 24 or 48 h with tested compounds at
4.1.4.1. (Z)-2-(6-Fluoro-2-methyl-3-(4-(methylsulfinyl)benzylidene)-
3H-inden-1-yl)-N (1). The title compound was prepared from 4
following the general procedure. The crude product was purified by
titration with Et2O to give 1 as yellow solid, mp: 208e209 ꢁC; yield
200 mM, harvested by trypsinization and combined with detached
cells. Apoptosis was then assessed by the Cell Death Detection
ELISA kit (Roche, Mannheim, Germany) based on the recognition of
released nucleosomes after DNA internucleosomal fragmentation
by a mouse monoclonal antibody directed against DNA and
histones. The assay allowed the specific determination of mono-
and oligo-nucleosomes in the cytoplasmic (apoptosis) or in the
supernatant fraction (necrosis) of cell lysates.
70%; 1H NMR (CDCl3,
d ppm): 7.76e7.64 (m, 4H); 7.21e7.15 (m, 2H);
6.87e6.81 (m, 1H); 6.65e6.57 (m, 1H); 3.58 (s, 2H); 2.82 (s, 3H);
2.22 (s, 3H). Anal. Calcd for C20H18FNO3S: C, 64.68; H, 4.88; N, 3.77.
Found: C, 64.87; H, 5.12; N, 3.96.
4.1.4.2. (Z)-2-(6-Fluoro-2-methyl-3-(4-(methylsulfonyl)benzylidene)-
3H-inden-1-yl)-N-hydroxyacetamide (2). The title compound was
prepared from 5 following the general procedure. The crude product
was purified by titration with Et2O to give 2 as yellow solid; mp:
4.2.5. RT-PCR assay
RNA from cells was extracted by using the RNeasy Mini kit and
reverse-transcribed by the QuantiTect Reverse Transcription kit.
PCR was performed by the Hot StartTaq Master Mix kit. Primers
used were: 50-GCCTGACTCCTTCAAGATCG-30 (F) and 50-AGGGA-
CAGGTCTTGGTGTTG-30 (R) for COX-1; 50-TGAAACCCACTCCAAA-
CACA-30 (F) and 50-AACTGATGCGTGAAGTGCTG-30 (R) for COX-2;
50-TCTGACGGCAACTTCAACTG-30 (F) and 50-TTGAGGAGTCTCACC-
CAACC-30 (R) for Bax; 50-TCCATGTCTTTGGACAACCA-30 (F) and
50-CTCCACCAGTGTTCCCATCT-30 (R) for Bcl-2; 50-GTGAAGGTCG-
GAGTCAACG-30 (F) and 50-GGTGAAGACGGCCAGTGGACTC-30 (R) for
GAPDH and the expected amplification products were 446, 385,
188, 203 and 300 bp long, respectively. Relative densitometry of
bands was measured using NIH ImageJ gel analysis.
210e215 ꢁC; yield 70%; 1H NMR (CDCl3,
d ppm): 7.97e7.93 (m, 2H);
7.72e7.68 (m, 2H); 7.19e7.10 (m, 2H); 6.94e6.83 (m, 1H); 3.60 (s,
2H); 3.11 (s, 3H); 2.22 (s, 3H). Anal. Calcd for C20H18FNO4S: C, 62.00;
H, 4.68; N, 3.62. Found: C, 61.83; H, 4.81; N, 3.75.
4.1.4.3. (Z)-2-(6-Fluoro-2-methyl-3-(4-(methylthio)benzylidene)-3H-
inden-1-yl)-N-hydroxyacetamide (3). The title compound was
prepared from 6 following the general procedure. The crude
product was purified by titration with Et2O to give 3 as yellow solid,
mp: 199e201 ꢁC; yield 70%; 1H NMR (CDCl3,
d ppm): 7.46e7.30 (m,
4H); 7.18 (s, 2H); 6.86e6.81 (m,1H); 6.65e6.57 (m,1H); 3.54 (s, 2H);
2.55 (s, 3H); 2.19 (s, 3H). Anal. Calcd for C20H18FNO2S: C, 67.59; H,
5.10; N, 3.94. Found: C, 67.73; H, 5.23; N, 4.12.
4.2.6. Anti-angiogenic activity
HUVEC were sub-cultured into twenty-four-well plates at
a density of 104 cells/well over night and kept in serum-free
medium for 24 h. Cells were then incubated in growth medium in
4.2. Pharmacology
the presence or absence of 0.1 mM test compounds for 96 h expo-
4.2.1. Cell culture conditions
The human cancer cell lines MIA PaCa-2 (pancreas) and COLO320
(colon) (American Type Culture Collection, Manassas, VA), were
sure. At the end of the experiments, cells were harvested and
counted by haemocytometer and changes in cell growth were
expressed as a percentage relative to day 0.
cultured in DMEM medium supplemented with
10% fetal bovine serum, 2.5% horse serum, 50 IU/ml penicillin and
50
g/ml streptomycin (SigmaeAldrich, Milano, Italy) at 37 ꢁC in an
atmosphere of 5% CO2. Human umbilical vein endothelial cells
(HUVEC) were cultured in an optimized media containing the
angiogenic factors, bFGF and VEGF (Lonza, Walkersville, MD, USA).
L-glutamine (2 mM),
4.2.7. Data analysis and statistical procedures
m
Data were expressed as mean ꢀ standard error of the mean
(SEM) of three measurements. Data analysis was performed with
GraphPad Prism software. The 50% inhibitory concentration of cell
growth (IC50) was calculated with a non-linear least squares curve

S. Fogli et al. / European Journal of Medicinal Chemistry 45 (2010) 5100e5107
5107
[12] S.S. Tam, D.H. Lee, E.Y. Wang, D.G. Munroe, C.Y. Lau, Tepoxalin, a novel dual
inhibitor of the prostaglandin-H synthase cyclooxygenase and peroxidase
activities, J. Biol. Chem. 270 (1995) 13948e13955.
[13] S.K. Seo, H.O. Jin, H.C. Lee, S.H. Woo, E.S. Kim, D.H. Yoo, S.J. Lee, S. An,
C.H. Rhee, S.I. Hong, T.B. Choe, I.C. Park, Combined effects of sulindac and
suberoylanilide hydroxamic acid on apoptosis induction in human lung
cancer cells, Mol. Pharmacol. 73 (2008) 1005e1012.
[14] S. Kikuchi, H. Konishi, Y. Hashimoto, The deoxygenation of sulfoxide mediated
by the Ph3P/lewis acid combination and the application to the kinetic reso-
lution of racemic phosphines using optically active sulfoxide, Tetrahedron 61
(2005) 3587e3591.
using a sigmoid doseeresponse (variable slope) equation. Statis-
tical analysis was carried out by one-way analysis of variance
(ANOVA) followed by the NewmaneKeuls test for multiple
comparison. P < 0.05 was taken as level of significance.
References
[1] D. Wang, R.N. Dubois, Eicosanoids and cancer, Nat. Rev. Cancer 10 (2010)
181e193.
[15] PCT No. WO 2007081694, Int. Appl. (2007). 47p.
[2] N.M. Davies, M.S. Watson, Clinical pharmacokinetics of sulindac. A dynamic
old drug, Clin. Pharmacokinet. 32 (1997) 437e459.
[3] M.J. Thun, S.J. Henley, C. Patrono, Nonsteroidal anti-inflammatory drugs as
anticancer agents: mechanistic, pharmacologic, and clinical issues, J. Natl.
Cancer Inst. 94 (2002) 252e266.
[4] H.P. Ciolino, S.E. Bass, C.J. MacDonald, R.Y. Cheng, G.C. Yeh, Sulindac and its
metabolites induce carcinogen metabolizing enzymes in human colon cancer
cells, Int. J. Cancer 122 (2008) 990e998.
[5] P.L. Rice, R.J. Goldberg, E.C. Ray, L.J. Driggers, D.J. Ahnen, Inhibition of extra-
cellular signal-regulated kinase 1/2 phosphorylation and induction of
apoptosis by sulindac metabolites, Cancer Res. 61 (2001) 1541e1547.
[6] C. Herrmann, C. Block, C. Geisen, K. Haas, C. Weber, G. Winde, T. Möröy,
O. Müller, Sulindac sulfide inhibits Ras signaling, Oncogene 17 (1998)
1769e1776.
[7] M.T. Taylor, K.R. Lawson, N.A. Ignatenko, S.E. Marek, D.E. Stringer, B.A. Skovan,
E.W. Gerner, Sulindac sulfone inhibits K-ras-dependent cyclooxygenase-2
expression in human colon cancer cells, Cancer Res. 60 (2000) 6607e6610.
[8] S. Flis, D. Soltysiak-Pawluczuk, A. Jedrych, Z. Jastrzebski, M. Remiszewska,
J. Splawinski, Antiangiogenic effect of sulindac sulfide could be secondary to
induction of apoptosis and cell cycle arrest, Anticancer Res. 26 (2006)
3033e3041.
[9] D.J. de Groot, E.G. de Vries, H.J. Groen, S. de Jong, Non-steroidal anti-inflam-
matory drugs to potentiate chemotherapy effects: from lab to clinic, Crit. Rev.
Oncol. Hematol. 61 (2007) 52e69.
[10] C.H. Hennekens, S. Borzak, Cyclooxygenase-2 inhibitors and most traditional
nonsteroidal anti-inflammatory drugs cause similar moderately increased risks
of cardiovascular disease, J. Cardiovasc. Pharmacol. Ther. 13 (2008) 41e50.
[11] I. Moench, H. Prentice, Z. Rickaway, H. Weissbach, Sulindac confers high level
ischemic protection to the heart through late preconditioning mechanisms,
Proc. Natl. Acad. Sci. U.S.A. 106 (2009) 19611e19616.
[16] A. Rossello, E. Nuti, E. Orlandini, P. Carelli, S. Rapposelli, M. Macchia,
F. Minutolo, L. Carbonaro, A. Albini, R. Benelli, G. Cercignani, G. Murphy,
A. Balsamo, Bioorg. Med. Chem. 12 (2004) 2441e2450.
[17] T.A. Chan, P.J. Morin, B. Vogelstein, K.W. Kinzler, Mechanisms underlying
nonsteroidal antiinflammatory drug-mediated apoptosis, Proc. Natl. Acad. Sci.
U.S.A. 95 (1998) 681e686.
[18] F.A. Sinicrope, R.C. Penington, Sulindac sulfide-induced apoptosis is enhanced
by a small-molecule Bcl-2 inhibitor and by TRAIL in human colon cancer cells
overexpressing Bcl-2, Mol. Cancer Ther. 4 (2005) 1475e1483.
[19] W.N. William Jr., J.V. Heymach, E.S. Kim, S.M. Lippman, Molecular targets for
cancer chemoprevention, Nat. Rev. Drug Discov. 8 (2009) 213e225.
[20] M.T. Yip-Schneider, D.S. Barnard, S.D. Billings, L. Cheng, D.K. Heilman, A. Lin,
S.J. Marshall, P.L. Crowell, M.S. Marshall, C.J. Sweeney, Cyclooxygenase-2
expression in human pancreatic adenocarcinomas, Carcinogenesis 21 (2000)
139e146.
[21] L. Yang, Z. Cao, H. Yan, W.C. Wood, Coexistence of high levels of apoptotic
signaling and inhibitor of apoptosis proteins in human tumor cells: implica-
tion for cancer specific therapy, Cancer Res. 63 (2003) 6815e6824.
[22] C.M. Ulrich, J. Bigler, J.D. Potter, Non-steroidal anti-inflammatory drugs for
cancer prevention: promise, perils and pharmacogenetics, Nat. Rev. Cancer 6
(2006) 130e140.
[23] L. Thon, H. Möhlig, S. Mathieu, A. Lange, E. Bulanova, S. Winoto-Morbach,
S. Schütze, S. Bulfone-Paus, D. Adam, Ceramide mediates caspase-indepen-
dent programmed cell death, FASEB J. 19 (2005) 1945e1956.
[24] S.E. Witta, D.L. Gustafson, A.S. Pierson, A. Menter, S.N. Holden, M. Basche,
M. Persky, C.L. O’Bryant, C. Zeng, A. Baron, M.E. Long, A. Gibbs, K. Kelly,
P.A. Bunn Jr., D.C. Chan, P. Pallansch, S.G. Eckhardt, A phase I and pharma-
cokinetic study of exisulind and docetaxel in patients with advanced solid
tumors, Clin. Cancer Res. 10 (2004) 7229e7237.