A Novel Sulindac Derivative Inhibits Lung AdenocarcinomaCell Growth through Suppression of Akt/mTORSignaling andInduction of Autophagy
Evrim Gurpinar1, William E. Grizzle2, John J. Shacka2, Burton J. Mader2, Nan Li3, Nicholas A. Piazza4,Suzanne Russo5, Adam B. Keeton5, and Gary A. Piazza5
AbstractNonsteroidal anti-inflammatory drugs such as sulindac sulfide have shown promising antineoplastic
activity in multiple tumor types, but toxicities resulting from COX inhibition limit their use in cancer therapy.
We recently described aN,N-dimethylethyl amine derivative of sulindac sulfide, sulindac sulfide amide (SSA),
that does not inhibit COX-1 or -2, yet displays potent tumor cell growth–inhibitory activity. Here, we studied
the basis for the growth-inhibitory effects of SSA on human lung adenocarcinoma cell lines. SSA potently
inhibited the growth of lung tumor cells with IC50 values of 2 to 5 mmol/L compared with 44 to 52 mmol/L for
sulindac sulfide. SSA also suppressed DNA synthesis and caused a G0–G1 cell-cycle arrest. SSA-induced cell
death was associated with characteristics of autophagy, but significant caspase activation or PARP cleavage
was not observed after treatment at its IC50 value. siRNA knockdown of Atg7 attenuated SSA-induced
autophagy and cell death,whereas pan-caspase inhibitor ZVADwas not able to rescue viability. SSA treatment
also inhibited Akt/mTOR signaling and the expression of downstream proteins that are regulated by this
pathway. Overexpression of a constitutively active form of Akt was able to reduce autophagy markers and
confer resistance to SSA-induced cell death. Our findings provide evidence that SSA inhibits lung tumor cell
growthbyamechanism involving autophagy induction through the suppression ofAkt/mTORsignaling. This
unique mechanism of action, along with its increased potency and lack of COX inhibition, supports the
development of SSA or related analogs for the prevention and/or treatment of lung cancer.Mol Cancer Ther;
12(5); 663–74. �2013 AACR.
IntroductionLung cancer is the leading cause of cancer death in
patients older than 40 years, accounting for 226,160 newcancers and 160,340 deaths annually in the United States(1). Lung cancer is usually aggressive and characterizedby early progression and metastases, as approximately60% of patients with small cell lung cancer (SCLC) and upto 40% of patients with non–small cell lung cancer(NSCLC) will present with stage IV disease at the timeof diagnosis. The 5-year survival rate for patients withlung cancer is only 15.9%, with most dying from metas-tasis (2). Early detection of lung cancer can improveoutcome as evident from the National Lung Screening
Trial (NLST, ACRINA6654) that showedmortality can bereduced by 20% from screening high-risk patients withlow-dose helical computed tomography compared withchest X-ray (3). Similarly, data from the InternationalEarly Lung Cancer Action Program (I-ELCAP) showedthat stage I lung cancer detected by annual low-dosecomputed tomography scans improved survival withearly treatment (4). With the potential for early detectionof lung cancer in high-risk populations, it is important todevelop effective new therapeutic strategies.
Epidemiological studies indicate that long-term useof nonsteroidal anti-inflammatory drugs (NSAID) is asso-ciated with a significant decrease in cancer incidence anddelayed progression in patients with colorectal, breast,lung, and other cancers (5, 6). The use of NSAIDs is alsoassociated with reduced risk from cancer-related mortal-ity and distant metastasis (7–9). With regard to lungcancer, NSAIDs have been shown to reduce the incidenceof lung carcinoma by 21%. The reduction in risk isincreased to 32% when adjusted for smoking, and theefficacy is greatest in former smokerswith a risk reductionof as much as 42% (10). Rodent studies support theseobservations by showing that NSAIDs inhibit tumor for-mation in carcinogen-inducedmodels of lung tumorigen-esis (11–13). Unfortunately, the long-term use of NSAIDs
Authors' Affiliations: Departments of 1Pharmacology and Toxicology,2Pathology, and 3Biochemistry, The University of Alabama at Birmingham;4The University of Alabama at Birmingham School of Medicine, Birming-ham; and 5Drug Discovery Research Center, Mitchell Cancer Institute,University of South Alabama, Mobile, Alabama
Corresponding Author: Gary A. Piazza, Mitchell Cancer Institute,University of South Alabama, 1660 Springhill Avenue, Suite 3029,Mobile, AL 36604. Phone: 251-445-8412; Fax: 251-460-6994; E-mail:[email protected]
and COX-2 inhibitors is not recommended because ofpotentially fatal gastrointestinal, renal, and cardiovascu-lar toxicities associated with the depletion of physiolog-ically important prostaglandins resulting from their COX-1 and -2 inhibitory activity (14, 15).
Because smoking contributes to chronic pulmonaryinflammation, it is logical that NSAIDs may inhibit lungcancer development through suppressing inflammatorymediators such as prostaglandins. However, some inves-tigators have implicated alternative mechanisms bywhich NSAIDs may exert their antineoplastic activity,suggesting that it may be feasible to develop derivativesthat do not inhibit COX-1 or -2 and have reduced toxicity(16–19). As support for a COX-independent mechanism,the sulfone metabolite of sulindac, which does not inhibitCOX-1 and -2, has been reported to inhibit tumorigenesisin colon, mammary, lung, and bladder models (20–22). Ina mouse model of tobacco carcinogen–induced lungtumorigenesis, sulindac sulfone strongly inhibited bothtumor incidence and multiplicity (13). Sulindac sulfonealso inhibited tumor growth and metastasis that led toincreased survival of rats with orthotopically implantedhuman lung tumors alone or in combination with doc-etaxel (23–25). Although sulindac sulfone did not receiveU.S. Food and Drug Administration approval for humanuse because of hepatotoxicity, these studies support thedevelopment of other non–COX-inhibitory derivatives ofsulindac.
We recently described a N,N-dimethylethyl aminederivative of sulindac sulfide referred to as sulindacsulfide amide (SSA) that does not inhibit COX-1 or -2, yetinhibits colon tumor cell growth in vitro and in vivo (26).Because of the strong efficacy of sulindac sulfone in lungcancermodels, we conducted additional studieswith SSAin human lung cancer cell lines to determine their level ofsensitivity and to investigate the underlying mechanismof action. Here, we describe a novel component of SSA-induced cytotoxicity that involves autophagy inductionvia suppression of Akt/mTOR signaling.
Materials and MethodsDrugs and reagents
SSA was synthesized and characterized as describedpreviously (26). Lipofectamine LTX and PLUS transfec-tion reagents were purchased from Invitrogen. LC3 anti-body was purchased from Novus Biologicals. Akt1/2/3(pan-Akt) antibody was from Santa Cruz Biotechnology,MDM2 antibody was from EMD Biosciences and p62antibody was from Abgent. All other antibodies werepurchased from Cell Signaling Technology. pEGFP-LC3and ptfLC3 plasmids were provided by Dr. John Shacka(University of Alabama at Birmingham, Birmingham,AL). Constitutively active Akt (Myr-Akt1, Addgene plas-mid 9008) and empty vector (pcDNA3, Addgene plasmid10792) plasmids were purchased fromAddgene. Z-VAD-FMK was purchased from EMD Chemicals. All otherdrugs and reagents were purchased from Sigma-Aldrichunless otherwise stated.All compoundsweredissolved in
dimethyl sulfoxide (DMSO), and the maximum finalconcentration of DMSO was 0.1% in all experiments.
Cell cultureThe human lung adenocarcinoma cell lines A549,
H1299, and HOP-62 were obtained from the AmericanType Culture Collection (ATCC) and grown under stan-dard cell culture conditions in RPMI-1640 containing 5%FBS at 37�C in a humidified atmosphere with 5% CO2. AllATCC cell lines were expanded upon delivery, andnumerous vials of low-passage cells were preserved inliquid N2. No vial of cells was passaged for more than 2months. Cell line characterization is conducted by ATCCthrough STR profiling and reauthentication was notconducted.
Cell viability assayTissue culturemicrotiter 96-well plateswere seeded at a
density of 5,000 cells per well and incubated for 18 to 24hours before being treated with the specified compoundor vehicle control. The inhibition of cell growth caused bytreatment was determined as described previously (27).
Fluor 488 & Propidium Iodide (PI) Dead Cell Apoptosiskit from Invitrogen. In brief, 2 � 105 to 3 � 105 cells wereexposed to SSA or vehicle control in 6-well plates andincubated for 24 hours before analysis. The cellswere thenharvested and analyzed with a Becton Dickinson FACS-Calibur instrument (excitation, 488 nm; emission, 530 nm)according to manufacturer’s instructions. The cells thatwere positive for Annexin V alone, and Annexin V and PIwere counted. Activity of caspase-3 and -7 was measuredusing theCaspase-Glo 3/7Assay (Promega) as previouslydescribed (27). PARP cleavage was measured byWesternblotting.
Cell proliferation assayCell proliferation was determined by using the Click-iT
EdU Alexa Fluor 488 Proliferation Assay (Invitrogen).Cells were seeded at a density of 1 � 106 cells per 10-cmtissue culture dish and incubated with SSA, sulindacsulfide, or vehicle control. Six hours after initial dosing,5-ethynyl-20-deoxyuridine (EdU, 10 mmol/L) was addedinto the cell culture media and cells were incubated for anadditional 18 hours. Cells were harvested and analyzedaccording to themanufacturer’s instructions. Thepercent-age of proliferating cells was quantified by using a BectonDickinson FACSCalibur instrument.
Cell-cycle measurementsCells (2� 105 to 3� 105) were exposed to SSA, sulindac
sulfide, or vehicle control in 6-well plates and incubatedfor 24 hours before analysis. The cells were trypsinized,washedwith PBS, and fixed in 1mL of 70% ethanol at 4�Covernight, followed by incubationwithRNase (1mg/mL)and staining with PI (40 mg/mL). DNA content was
Gurpinar et al.
Mol Cancer Ther; 12(5) May 2013 Molecular Cancer Therapeutics664
determined by flow cytometry using a Becton DickinsonFACSCalibur instrument.
Visualization of intracellular vacuolesAcidic compartments were visualized by labeling the
cells with Lyso-IDGreen detection reagent (Enzo) accord-ing to the manufacturer’s instructions. Nuclei were coun-terstained with DRAQ5. Images were obtained using anEvotec Opera confocal microscope with a �20 objectivelens. For histologic analysis, cells were harvested byCellStripper (CellGro), resuspended in PBS and smearedon microscope slides. Cells were then immediately fixedin 10% neutral-buffered formalin before being stainedwith hematoxylin and eosin.
Transmission electron microscopyCells were seeded at a density of 1� 106 cells per 10-cm
cell culture dish and treated with SSA or vehicle controlfor 24 hours. Cells were harvested using CellStripper(CellGro), washed once in PBS, and then fixed with 2%glutaraldehyde in 0.2 mol/L HEPES, pH 7.4, at roomtemperature for 2 hours. Additional sample processingand transmission electron microscopy (TEM) were con-ducted as described previously by Shacka and colleagues(28).
TransfectionsAll plasmid transfections were conducted in OptiMEM
reduced-serum medium (Invitrogen) containing 0.3%Lipofectamine LTX transfection reagent, 0.1% PLUSreagent, and 1 mg/mL DNA. Cells were incubated for24 hours before treatment. For RNA interference experi-ments, A549 cells (75% confluent) were transfected byusing the Amaxa Nucleofector II device protocol (Lonza).Briefly, 1� 106 cells were pelleted and resuspended in 100mL of Nucleofactor reagent followed by addition of eithercontrol or ATG7 siRNA (200 nmol/L). Transfer of thereaction mixture was completed by electroporation in theAmaxa Nucleofector II Device. X-001 Nucleofector pro-gram was used. After transfection, cells were transferredto culture plates for 18 hours before experiments were setup. siRNAs corresponding to the human cDNA sequencefor ATG7 and the nonsilencing negative control siRNAwere from Dharmacon Research.
Autophagic imaging and flux assaysCells seeded on coverslips at 50% confluency were
transiently transfected with either the eGFP-LC3 or thetfLC3 plasmid. Twenty-four hours after transfection, cellculture media were replaced with fresh and cells weretreated with 5 mmol/L SSA or vehicle control. After 24hours of treatment, cells were fixed in 10% neutral-buff-ered formalin for 10minutes, then rinsedwith Tris buffer,and counterstained with 40,6-diamidino-2-phenylindole(DAPI). Slides were mounted using Fluoromount (Sig-ma). Imaging was conducted on a Zeiss Axio Imager.M2fluorescence microscope connected to a Zeiss AxioCamMRm camera. Autophagy induction was quantified by
counting the number of cellswith eGFP-LC3 translocationintodots (aminimumof 700 cells/sample) after treatment.Image processing was conducted by using the NIH Ima-geJ software.
Western blot analysisImmunoblottingwas conducted as described previous-
ly (27). For p-mTOR, a 5% PAGE was used to achieveproper separation.
Experimental design and data analysisDrug effects on cell growth and IC50 values were deter-
mined as described previously (27). All experiments wererepeated a minimum of 3 times to determine the repro-ducibility of the results.All values represent a comparisonbetweendrug treatment at the specified concentration andvehicle-treated controls. All error bars represent SEM.Statistical analysis was conducted using Student t testwith P < 0.05 deemed as statistically significant.
ResultsIn vitro tumor cell growth–inhibitory activity of SSA
Previous molecular modeling studies revealed theimportance of the carboxylic acid group on sulindacsulfide for inhibition of COX-1 and -2 (26). As depictedin Fig. 1A, SSA was designed by substituting the carbox-ylic acid on sulindac sulfide for a N,N-dimethylethylaminemoiety to block COXbinding. Despite lacking COXinhibitory activity, SSA was found to potently inhibit thegrowth of human lung adenocarcinoma cell lines, A549,H1299, and HOP-62 with IC50 values ranging from 2 to 5mmol/L (Fig. 1B). By comparison, the nonselective COX-1/2 inhibitor, sulindac sulfide,was approximately 10 to 20times less potent with IC50 values ranging from 44 to 52mmol/L. Next, we examined whether the growth-inhib-itory effects of SSA involved apoptosis induction and/orinhibition of proliferation. To assess the induction ofapoptosis, cells were treated with SSA for 24 hours andthe activation of effector caspase-3 and -7, which arespecific biochemical markers of apoptotic cell death, wasmeasured. As shown in Fig. 1C, SSA treatment was notable to induce significant caspase activation at its IC50
value in either cell type. However, caspase activation bySSAwasdose-dependent, andhigher concentrationswereable to lead to significant levels of apoptosis comparedwith vehicle control. This activation, nonetheless, waslow compared with that induced by sulindac sulfideand the apoptosis-inducing agent staurosporine. We alsoassessed PARP cleavage, which occurs downstream ofcaspase activation, as an additional specific marker ofapoptosis by Western blotting. We were able to detectcleavedPARP after 10mmol/LSSA treatment but not after5 mmol/L treatment (Fig. 1D). A dose-dependent increasein PARP cleavagewas also observed after sulindac sulfidetreatment. As an additional marker of cell death, wemeasured Annexin-V and PI labeling after SSA treatmentby flow cytometry. Interestingly, we observed a signifi-cant and dose-dependent increase in the extent of
Novel Sulindac Analog Induces Autophagy in Lung Cancer Cells
www.aacrjournals.org Mol Cancer Ther; 12(5) May 2013 665
Annexin-V surface staining and PI labeling in all cell linesafter 24 hours of SSA treatment at concentrations thatsuppressed growth and lower than those required forcaspase activation (Fig. 1E). SSA was also able to signif-icantly and dose dependently inhibit DNA synthesis inlung tumor cells as measured by EdU incorporation (Fig.1F). Furthermore, SSA treatment led to a dose-dependentincrease in thepercentage of cells in theG0–G1phase of thecell cycle at concentrations that suppressed growth (Fig.1G). Significant inhibition of DNA synthesis and G0–G1
cell-cycle arrest were also observed after single-dosesulindac sulfide treatment at 100 mmol/L. These resultsindicate that the growth-inhibitory activity of SSA isassociated with inhibition of proliferation but not apopto-
sis after treatment at its IC50 value, suggesting the involve-ment of alternative mechanisms of cell death.
SSA induces autophagy in lung adenocarcinoma cellsSSA-treated cells displayed a striking accumulation of
vesicle-like structures within the cytoplasm that wasreadily apparent by phase-contrast microscopy (Fig. 2A,left). As shown by hematoxylin and eosin staining (Fig.2A, middle), SSA treatment resulted in extensive intra-cellular vacuolization characterized by multiple smallvesicles (arrowheads) and large vacuoles (arrow) typical-ly observed in cells undergoing autophagy. To determinewhether SSA induced an autophagic response in lungcancer cells, we used the Lyso-ID dye, which labels acidic
Sulindac sulfide
(SS)
FF
S
A549
H1299
A549
A549
H1299
H1299
HOP62
HOP62
A549H1299HOP62
G0–G1
G2–MS
*,P < 0.05
*,P < 0.05
**,P < 0.01
*,P < 0.05
**,P < 0.01
*,P < 0.05**,P < 0.01
S
OHHN
A B
C D
E
G
F
N
50
0
0.0 2.5 5.0
% G
ate
d c
ells
% G
ate
d c
ells
% G
ate
d c
ells
7.5
+ 2.5 5
SSA
SSA (µmol/L)
SSA (µmol/L)0 2.5 5 7.5 SS
SSA (µmol/L)
0 2.5 5 7.5 SS
SSA (µmol/L)0 2.5 5 7.5 SS
SSA (µmol/L)0 2.5 5 7.5 SS
SSA (µmol/L)
SS (µmol/L)
SS
7.5 10 100
0
Pro-PARP
Cleaved-PARP
GAPDH
Pro-PARP
Cleaved-PARP
GAPDH
+ 5 10
0 + 60 90
(µmol/L)
Ca
sp
ase
-3/7
activity
(fo
ld in
cre
ase
)
30
80
% C
ell
death
60
40
20
0
20
10
6
4
2
0
O
100
80
60
40
20
0
0.1 1
HOP62, SSA, IC50
= 1.95 µmol/L
HOP62, SS, IC50
= 51.80 µmol/L
H1299, SSA, IC50
= 3.05 µmol/L
H1299, SS, IC50
= 48.57 µmol/L
A549, SSA, IC50
= 4.72 µmol/L
A549, SS, IC50
= 44.00 µmol/L
10
Concentration (µmol/L)
100
% G
row
th inhib
itio
n
(com
pare
d w
ith c
ontr
ol, 7
2 h
)
100
80
60
40
20
0
100
80
60
40
20
0
100
80
60
40
20
0
80
60
40
20
0
% P
rolif
era
tion
Sulindac sulfide amide
(SSA)
O
Figure 1. Inhibition of growth,induction of apoptosis, andinhibition of cell proliferation inlung adenocarcinoma cells bySSA. A, structures of sulindacsulfide (SS) and SSA. B, dose-dependent growth-inhibitoryactivity of SS and SSA in A549,H1299, and HOP-62 cells after72 hours of treatment.C, apoptosis induction asmeasured by caspase-3 and -7activation. D, PARP cleavage after24 hours of treatment with SSAand SS (100 mmol/L; þ, 1 mmol/Lstaurosporine). E, Annexin V/PIlabeling after 24 hours of SSAtreatment. F, dose-dependentinhibition of DNA synthesis asmeasured by EdU incorporation.G, induction of G0–G1 cell-cyclearrest in A549, H1299, and HOP-62 cells after 24 hours of treatmentwith SSA or SS (100 mmol/L).
Gurpinar et al.
Mol Cancer Ther; 12(5) May 2013 Molecular Cancer Therapeutics666
Figure 2. SSA induces autophagy in lung adenocarcinoma cells. A, phase-contrast images (left) and acidic vesicle-specific dye labeling (right) of A549 cellstreatedwith vehicle or 5 mmol/L SSA for 24 hours. Hematoxylin and eosin staining of floating cells after 10 mmol/L SSA treatment for 24 hours (middle). Arrowsindicate intracellular vesicles. B, eGFP-LC3 imaging of A549 and H1299 cells after vehicle or 5 mmol/L SSA treatment for 24 hours. Autophagic vacuoles areidentified as punctuate dots within the cytoplasm. Scale bars, 10 mm. C, dose- and time-dependent induction of autophagy in A549 and H1299cells as indicated by LC3-I to LC3-II conversion. Cells were treated either with the indicated concentrations of SSA for 24 hours or with 5 mmol/L SSA for theindicated time periods and subjected to Western blotting. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as loading control. Resultsshown are representative of 2 independent experiments. D, quantification of eGFP-LC3–positive puncta in A549 and H1299 cells after 24 hours of SSAtreatment. E, representative electron micrographs of A549 cells treated with vehicle or 7.5 mmol/L SSA for 24 hours. Nuclei are labeled N. Arrowsindicate autophagic vacuoles. Scale bars, 2 mm. F, magnified image shows detailed autophagic vacuoles structure and residual digested cellular materialwithin their lumen. Scale bars, 500 nm. G, representative electron micrographs of A549 cells at different stages of autophagy after 7.5 mmol/L SSA treatmentfor 24 hours. Images show progressive autophagic degradation (clockwise) and an absence of morphological features of apoptosis. Scale bars, 2 mm.
Novel Sulindac Analog Induces Autophagy in Lung Cancer Cells
www.aacrjournals.org Mol Cancer Ther; 12(5) May 2013 667
compartments within the cytoplasm including autopha-gosomes. SSA treatment (5 mmol/L, 24 hours) of A549cells increased both the number and size of acidic vesicleswithin the cytoplasm compared with vehicle treatment(Fig. 2A, right). To confirm that the acidic vesicles repre-sented autophagosomes, A549 cells were transfectedwitha plasmid that expresses GFP-tagged microtubule-asso-ciated protein 1 light-chain 3 (LC3), a cytoplasmic protein,which, at the onset of autophagy, is cleaved, lipidated, andincorporated into the membranes of autophagosomes(29). GFP-LC3 can be used to visualize and quantifyautophagosomes, which appear as punctate dots withinthe cytoplasm. After SSA treatment (5 mmol/L, 24 hours),the number of LC3-positive vesicles was significantlyincreased compared with vehicle control in both A549and H1299 cells, confirming that SSA-treated cells wereundergoing autophagy (Figs. 2B and 2D). SSA treatmentalso induced dose- and time-dependent cleavage of cyto-solic LC3-I to autophagosome-associated LC3-II (Fig. 2C).The concentrations required to induce autophagy werecomparable to the concentrations required for the inhibi-tion of growth and proliferation, suggesting that autop-hagic vacuole accumulation may contribute to the tumorcell growth–inhibitory activity of SSA.
To determine whether autophagy is contributing toSSA-induced cytotoxicity, the morphologic changesinduced by SSA treatment in A549 cells were examinedusing electron microscopy. Cells treated with SSA (7.5mmol/L, 24 hours) displayed extensive intracellularvacuolizationwithmultiple vesicles, distorted nuclei, anda reduced number of mitochondria (Fig. 2E). Consistentwith autophagy, multiple small autophagosomes distin-guished by their characteristic double-membranes couldbe identified (arrowheads), as well as large single-mem-braned vacuoles (arrows) containing electron-densedeposits, indicating mature autophagosomes. Highermagnification images clearly show that these autophagicvacuoles containeddegrading organelles (Fig. 2F). Similarultrastructural features were also observed in H1299 cellstreated with SSA (data not shown).
Figure 2G shows additional TEM images that capturethe progression of autophagic degradation in A549 cellsfrom early to advanced stages (clockwise). These imagesshow a dynamic process from the induction stage withmultiple smaller vesicles (top left), to the engulfmentof organelles (arrowheads), docking and fusion ofmature autophagosomes (arrows) to large autophago-lysosomes, and finally the complete degradation ofcytoplasmic material (bottom left). Remarkably, evenat the most advanced stages, this intense vacuolizationwas not paralleled by classical signs of apoptosis such asnuclear fragmentation, cell shrinkage, or chromatincondensation. These results indicate that SSA is able toinduce a distinct type of cell death involving autophagywith an absence of morphologic features of apoptosis.Furthermore, the Annexin V labeling results are con-sistent with previous reports indicating that phospha-tidylserine exposure on the cell surface may be a
common feature of apoptotic and autophagy-associatedcell death (30, 31).
SSA increases autophagic fluxTheaccumulation of autophagic vacuoles in response to
SSA treatment could result from either an increase in thesynthesis of autophagosomes or from a blockage of lyso-somal fusion and degradation. Elevated autophagic flux,whereby there is both active synthesis and degradation ofautophagasomes, has to be shown in cells undergoingautophagy-mediated cell death (32). To distinguish theseeffects, A549 cells were incubatedwith SSA and bafilomy-cin A1, a specific inhibitor of vacuolar type Hþ-ATPasethat blocks the last step of autophagic degradation (33).Asshown in Fig. 3A, LC3-II levelswere higher in cells treatedwith the combination of drugs compared with cells trea-ted with SSA only, indicating that SSA induces the syn-thesis of autophagosomes instead of blocking degrada-tion. The 20 nmol/L dose of bafilomycin A1was includedto show that complete blockage of lysosomal fusion wasachieved at the 10 nmol/L concentration used for thecombination treatment, and higher concentrations wereunable to induce further autophagic vacuoles accumula-tion. Disappearance of the ubiquitin-binding factor p62(SQSTM1), which targets polyubiquitinated proteins toautophagosomes for degradation, can also be used as amarker of increased autophagic turnover (34). Westernblotting showed a time-dependent decrease of p62 afterSSA treatment indicating that autophagic flux is increased(Fig. 3B).
The tfLC3 imaging assay was used as an additionalmarker of increased autophagic flux (35, 36). Overexpres-sion of the tfLC3 plasmid results in tandem expressionof both mRFP-LC3 and GFP-LC3. When autophago-somes fuse with the lysosomes, GFP-LC3 fluorescence isquenched because of the acidic environment. However,a compromise in lysosomal function stabilizes GFP-LC3fluorescence leading to GFP-LC3 and mRFP-LC3 colocali-zation. Accordingly, earlier autophagosomes would showcolocalization of GFP andmRFP but have diminished GFPas they mature and become acidified. Images from cellstreated with vehicle show one cell that has lost most ofitsGFPandasecondcell thatdisplaysGFP-LC3andmRFP-LC3 colocalization. While GFP-LC3 and mRFP-LC3–posi-tive vesicles are apparent after SSA treatment (5 mmol/L,24 hours), many others that have lost GFP (arrows) andcan be viewed as fusing with the lysosomes are observed(Fig. 3C). These findings indicate that autophagosomes inSSA-treated cells are able to fuse with the lysosomes andthat autophagic flux or turnover is increased.
Autophagy contributes to SSA-induced growthinhibition
The ability of SSA to induce autophagy in lung tumorcells led us to next investigate the precise role of autop-hagy in its anticancer activity. Cancer cells, including lungcancer cells, can undergo autophagy in response to var-ious anticancer therapies. This autophagic response can
Gurpinar et al.
Mol Cancer Ther; 12(5) May 2013 Molecular Cancer Therapeutics668
serve as a survival mechanism for the cell, or conversely,in some cases, as a nonapoptotic mechanism of pro-grammedcell death (37). Todeterminewhether the autop-hagy seen in SSA-treated lung cancer cells is involved incell death, we used siRNA to knockdown Atg7, an essen-tial protein for the induction of autophagy (38). Atg7siRNA was able to attenuate SSA-induced autophagy inA549 cells compared with scramble siRNA–transfectedcontrols, as measured by Western blotting (Fig. 4A).Furthermore, knockdown of Atg7 resulted in a significantincrease in cell viability after SSA treatment for 24 hours(Fig. 4B). To assess the role of apoptotic cell death in SSA-treated lung cancer cells,weusedpan-caspase inhibitorZ-VAD-FMK (40 mmol/L) to block caspase activity in A549andH1299 cells. Blocking caspase activationdidnot resultin an increase in viability in either cell line after SSAtreatment (Fig. 4C). In contrast, addition of Z-VAD-FMKwas able to reduce cytotoxicity induced by sulindac sul-fide and known apoptosis-inducing anticancer drug eto-poside in both cell lines. These results suggest that thegrowth-inhibitory activity of SSA is predominantlymedi-ated through autophagy and that cell death can occurindependent of apoptosis.
SSA modulates autophagy and loss of viability, inpart, by inhibiting Akt/mTOR/p70S6k signalingThe Akt/mTOR pathway, a known regulator of autop-
hagy and apoptosis, has been reported to be a target forsulindac sulfide (39, 40), as well as the selective COX-2inhibitor celecoxib (41, 42). In addition, several chemo-therapeutic agents have been shown to induce autophagyby inhibiting Akt and mTOR kinases (43). Therefore, weevaluated the effects of SSA treatment on the phosphor-
ylation of these kinases. As shown in Fig. 5A, SSA inhib-ited the phosphorylation of Akt at the Ser 473 residue in adose- and time-dependent manner without affecting totalAkt levels in both A549 and H1299 cells. The inhibition ofAkt phosphorylation became apparent starting at 4 hoursand reached maximal levels between 8 and 16 hours ofSSA treatment. In the same samples, we also probed forthe effects of SSA on mTOR kinase activity by measuringribosomal protein p70S6 kinase (p70S6k) phosphoryla-tion, a well-known mTOR substrate that is preferentiallyphosphorylated bymTOR at the Thr389 residue (44). SSAwas able to potently suppress p70S6 kinase phosphory-lation at the Thr389 in both cell lines (Fig. 5A). Theinhibitory effect was apparent after only 2 hours of treat-ment inA549 andH1299 cells, indicating a relatively rapidsuppression of mTOR kinase activity. Similar effects onAkt/mTOR signaling were also observed in HOP-62 cells(data not shown). To determine whether the inhibition ofAkt/mTOR signaling by SSA was involved in autophagyinduction and growth inhibition, lung cancer cells weretransfected with a plasmid expressing the myristoylatedand constitutively active form of the Akt1 protein (Myr-Akt; ref. 45). Overexpression of activated Akt was able toprevent the decrease in phosphorylated Akt levels andpartially block SSA-induced autophagy in A549 cells (Fig.5B).Weobtained similar resultswithH1299 cells (data notshown). Furthermore, the addition of Myr-Akt1 signifi-cantly inhibited SSA-induced cytotoxicity in both celllines (Fig. 5C). These results suggest that the inductionof autophagy and cell death by SSA is mediated throughthe Akt pathway. However, it needs to be noted that theblockage of autophagy and the rescue of viability areincomplete and additional mechanisms may be involved.
Figure 3. SSA increasesautophagic flux. A, LC3-IIaccumulation in A549 cells treatedwith 5 mmol/L SSA with or without10 nmol/L bafilomycin A1 (Baf.A1).Cells were pretreated with Baf.A1 for1 hour and then incubated with SSAfor an additional 4 hours. LC3-IIlevels reached a ceiling effect at10 nmol/L and higher concentrations(20 nmol/L) did not further increaseLC3-II levels. B, time-dependent p62downregulation after 5 mmol/LSSA treatment in A549 cells. C,effects of SSA on autophagic fluxwere measured using the ptfLC3plasmid that simultaneouslyexpresses mRFP- and GFP-taggedLC3 protein. A549 cells were treatedwith vehicle or 5 mmol/L SSA for 24hours. Early autophagosomes showGFP-mRFP colocalization whereaslate, acidic autophagosomes(arrows) lose GFP and appear red.Scale bars, 10 mm.
mRFP-LC3 eGFP-LC3 Merge
A549
A
C
BSSA (µmol/L)
Baf.A1 (nmol/L)
SS
A
Vehic
le
– – –5 5
– 10 20– 10
GAPDH
LC3-ILC3-II
GAPDH
p62
Time (h) Veh.(24 h) 2 4 8 16 24
Novel Sulindac Analog Induces Autophagy in Lung Cancer Cells
www.aacrjournals.org Mol Cancer Ther; 12(5) May 2013 669
SSA is a more potent and effective inducer ofautophagy than sulindac sulfide
We also evaluated sulindac sulfide for its ability toinduce autophagy and inhibit Akt/mTOR signaling.Figure 6A shows that while SSA induced autophagy atconcentrations equivalent to its IC50 value for growthinhibition (5 mmol/L), sulindac sulfide was unable toinduce autophagy at its IC50 value of 50 mmol/L inA549 cells as determined by LC3-I to LC3-II conversion.However, sulindac sulfide caused LC3-II accumulationafter 100 mmol/L treatment. These results indicate thatinduction of autophagy is a common feature amongsulindac sulfide and SSA but that higher concentrationsare needed in the case of sulindac sulfide,which exceed itsIC50 value for growth inhibition.
As shown in Fig. 6B, 100 mmol/L sulindac sulfidetreatment resulted in an inhibition of Akt phosphoryla-tion comparable to 2.5 mmol/L SSA treatment. Phosphor-ylation of mTOR kinase as well as its effector p70S6k wasalso inhibited by sulindac sulfide treatment. Anotherwell-known effector protein of mTOR kinase is the trans-lation repressor protein 4EBP-1. SSA and sulindac sulfidewere able to reduce the inactivating hyperphosphoryla-tion of 4E-BP1 indicating that both compounds inhibit
global protein translation, a feature of autophagy. Amoredetailed examination of other regulators of this pathwayshowed that phosphorylation of AMPK, which sup-presses mTOR activity by activating the TSC1/2 complex(46) was induced upon SSA treatment. This induction inAMPK activity was closely mimicked by 100 mmol/Lsulindac sulfide treatment.
In addition, SSA inhibited the phosphorylation ofMDM2 oncoprotein at the Akt-specific Ser 166 residue,which increases its interaction with p300 allowingMDM2-mediated ubiquitination and degradation ofp53 (47). However, the effect of sulindac sulfide onMDM2 phosphorylation was minimal at the 100mmol/L concentration. Finally, levels of the anti-apo-ptotic and autophagy-inhibiting protein, survivin, weremeasured following SSA and sulindac sulfide treat-ment. We and others have reported that sulindac sulfidecan suppress survivin levels leading to apoptosis and insome cases autophagic cell death (27, 48). We found thatSSA and sulindac sulfide potently suppress survivinlevels in A549 lung tumor cells. These results indicatethat SSA is a potent inhibitor of Akt and mTOR effectorsand suggest that these activities contribute to its anti-neoplastic activity.
A549 H1299
– ZVAD
+ ZVAD
**,P < 0.01
*,P < 0.05
**,P < 0.01
*,P < 0.05
A
C
B100
80
60
40
20
0
0
1 2.2 0.6 1.1
5
siCtrlsiCtrl
siAtg7
siAtg70 5SSA (µmol/L)
Atg7
SSA (µmol/L)
SSA (µmol/L)
% G
row
th
(com
pare
d w
ith c
ontr
ol)
100
80
60
40
20
% G
row
th
(com
pare
d w
ith c
ontr
ol) 100
80
60
40
20
% G
row
th
(com
pare
d w
ith c
ontr
ol)
0 2.5 5 7.5 10
0 2.5 5 7.5 SS Etop.
SSA (µmol/L)
0 2.5 5 7.5 SS Etop.
GAPDH
RatioLC3-II/GAPDH
LC3-I
LC3-II
Figure 4. Autophagymediates SSA-induced cancer cell death. A, modulation of SSA-induced autophagy by knockdown of Atg7. A549 cells were transfectedwith either Atg7 siRNA or a nonsilencing control siRNA for 24 hours and then treated with 5 mmol/L SSA for 4 hours. Levels of Atg7 and LC3-II weremeasured by Western blotting. B, knockdown of autophagy can attenuate SSA-induced cytotoxicity. A549 cells were transfected with either Atg7 or controlsiRNA for 24 hours and then treated with SSA for an additional 24 hours before measuring viability by CTG assay. C, inhibition of apoptosis does notlead to an increase in viability after SSA treatment. A549 and H1299 cells were pretreated with 40 mmol/L of pan-caspase inhibitor Z-VAD-FMK for 1 hour andthen further incubated with different doses of SSA, SS (100 mmol/L), and etoposide (100 mmol/L).
Gurpinar et al.
Mol Cancer Ther; 12(5) May 2013 Molecular Cancer Therapeutics670
DiscussionPreclinical, clinical, and epidemiologic studies have
shown promising antineoplastic properties for a num-ber of NSAIDs. Unfortunately, toxicity resulting fromCOX-1 and/or -2 inhibition limits their long-term usefor cancer indications. However, several lines of evi-dence suggest that COX-independent mechanisms maycontribute to or be fully responsible for their antineo-plastic activity. Perhaps the most compelling evidencefor this possibility comes from studies showing that thedose of a given NSAID to suppress tumor cell growthin vitro or in vivo far exceeds those required to inhibitCOX-1 or -2 (49, 50). Moreover, NSAID analogs ormetabolites that lack COX-inhibitory activity retain orhave improved tumor cell growth–inhibitory activity(17, 26, 51). Our results with SSA suggest that it is
feasible to chemically modify sulindac for developingsafer and more efficacious drugs for lung cancer che-moprevention or therapy.
Despite lacking COX-inhibitory activity, SSA was ableto inhibit lung tumor cell growth much more potentlythan sulindac sulfide. The increase in potency was appre-ciable with a shift from IC50 values of 44 to 52 mmol/L forsulindac sulfide to IC50 values of 2 to 5 mmol/L for SSA.Although further studies are necessary to determinewhether SSA can inhibit lung tumor growth in suitablepreclinical models and is sufficiently safe for human use,previous studies using a human colon xenograft mousemodel showed in vivo antitumor efficacy at dosages thatappeared to be well-tolerated (26).
Similar to previous results with sulindac sulfide,the tumor cell growth–inhibitory activity of SSA was
Figure 5. SSA modulates autophagy and cell death by inhibiting Akt/mTOR/p70S6k signaling. A, dose- and time-dependent downregulation of Akt andp70S6k phosphorylation after SSA treatment. Cells were treated either with the indicated concentration of SSA for 24 hours or with 5 mmol/L SSA forindicated time periods and subjected to Western blotting. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as loading control.Images are representative of 2 separate experiments. B, inhibition of SSA-induced autophagy by the overexpression of a constitutively active form ofAkt (Myr-Akt). Cells were transfectedwith an Akt overexpression (Myr-Akt) or empty vector (pcDNA3) plasmid for 24 hours and then treatedwith 5 mmol/L SSAfor 4 hours. Levels of phosphorylated Akt, total Akt, and LC3-II were measured by Western blotting. C, inhibition of SSA-induced cell death by Aktovexpression in A549 and H1299 cells. Cells were transfected with Myr-Akt or pcDNA3 plasmids for 24 hours and then treated with SSA for an additional24 hours before measuring viability by CTG assay.
Novel Sulindac Analog Induces Autophagy in Lung Cancer Cells
www.aacrjournals.org Mol Cancer Ther; 12(5) May 2013 671
associated with the inhibition of DNA synthesis andinduction of cell-cycle arrest. The ability of sulindac sul-fide to induce apoptosis has been shown in multiple celltypes and is generally associated with the antineoplasticproperties of sulindac and other NSAIDs (51–53). How-ever, in contrast to sulindac sulfide, appreciably lowerconcentrations of SSA equivalent to its IC50 value forgrowth inhibition caused a distinct type of cell death inlung adenocarcinoma cells involving the surface exposureof phosphatidylserine, as evident by Annexin V labeling,but in the absence of significant caspase cleavage or theclassical morphologic features of apoptosis such as nucle-ar fragmentation. Instead, SSA-induced cell death wascharacterized by extensive intracellular vacuolization andbulk degradation of cytoplasmic material with an appar-ently intact nucleus even at late stages. The intracellular
vacuoles induced by SSA treatment were confirmed to beautophagic vacuoles due to their acidic nature, translo-cation of autophagosome-associated LC3-II protein intotheir membranes, and the presence of digested organellematerial within their lumen as shown by TEM. Further-more, we showed that the increase in autophagic markerswas due to an increase in autophagic flux and not asso-ciated with the inhibition of lysosomal function. Weexplored the role of apoptosis and autophagic cell deathby using specific caspase inhibitors and knockdown ofautophagy through siRNA studies. Our results show thatautophagy but not apoptosis participates in SSA-inducedloss of cell viability at or around its IC50 value for growthinhibition. Although sulindac sulfide also induced autop-hagy, higher concentrations were required comparedwith concentrations required to suppress tumor cellgrowth.
On the basis of growing evidence suggesting that Akt/mTOR signaling can inhibit drug-induced autophagy, weexplored the role of this pathway on SSA-induced autop-hagy and cell death. SSA was able to potently inhibit theAkt/mTOR/p70S6k pathway at or around its IC50 valuefor growth inhibition. We were able to show, through theuse of a constitutively active Akt overexpression vector,that Akt signaling mediates SSA-induced autophagy andrelated cytotoxicity. We observed the inhibition of Akt/mTOR signaling after sulindac sulfide treatment as well,albeit at higher concentrations. These results suggestthat SSA and sulindac sulfide may share a similar mech-anismof action that impinges uponAkt/mTOR signaling.It is therefore plausible that the improved potency ofSSA to inhibit Akt signaling may underlie its improvedpotency to induce autophagy and cell-cycle arrest andhence inhibit growth by dual mechanisms involvingincreased tumor cell death and the suppression of prolif-eration. In contrast, the predominant mechanism of celldeath after sulindac sulfide treatment appears to beapoptosis.
The role of autophagy in tumor suppression is complexand likely involves several, sometimes paradoxical func-tions. Established tumors may use autophagy as an adap-tive response against metabolic stress such as starvation,hypoxia, oxidative damage, or chemotherapy (54). Incontrast, numerous researchers have shown using in vitroand in vivomodels that in response to certain chemother-apeutic agents, autophagy can alsomediate cell death as agenuine effector mechanism (55, 56). This has beendubbed autophagic cell death or programmed cell deathtype II. Alternatively, autophagy has been shown to act asan initial response mechanism for other agents, subse-quently triggering apoptotic events (57). As autophagyand apoptosis share some common effectors (e.g., Akt,Bcl-2, mTOR), it has been proposed that the nature andintensity of the initial stimulusmaydeterminewhich formof cell death will arise. In certain contexts of apoptosisresistance, for example, autophagy has been shown tomediate cell death in response to agents that wouldotherwise lead to apoptosis in na€�ve cells (58). This
0 2.5 5
5 µmol/L
SSA
50 µmol/L
SS
5 µmol/L
SSA
100 µmol/L
SS
Concentration (µmol/L)
SSA SS
7.5 100
GAPDH
p-AktSer473
p-mTORSer2448
p-MDM2Ser166
MDM2
Survivin
mTOR
Akt
p-AMPKThr172
AMPK
p-p70S6kThr389
p-4E-BP1Thr37/Thr46
4E-BP1
p70S6k
GAPDH
Veh. Veh.
A
B
LC3-ITreatment:
LC3-II
Figure 6. SSA inhibits the Akt/mTOR pathway and induces autophagymore potently than parent compound sulindac sulfide. A, LC3-II levelsafter 50 and 100 mmol/L SS treatment in A549 cells. Cells were treated for24 hours. SSA (5 mmol/L) was included as a control and for comparison.Autophagy activation can be seen after dosing cells with 100 mmol/L SS.B, effects of SSA and SS on additional mediators of autophagy signaling.A549 cells were treated at the indicated concentrations of SSA or100 mmol/L SS for 24 hours. SSA was able to dose dependently inhibitAkt, mTOR, p70S6k, 4E-BP1, and MDM2 phosphorylation. SSAtreatment induced AMPK phosphorylation while downregulating survivinlevels. SS was able to mimic SSA effects, albeit with low potency.
Gurpinar et al.
Mol Cancer Ther; 12(5) May 2013 Molecular Cancer Therapeutics672
suggests that autophagy can function as a distinct celldeath modality, which can be exploited for novel anti-cancer drug strategies to circumvent resistance.The ability of SSA to induce autophagy has important
implications for its development as a therapeutic agent forlung cancer, alone or in combination with standard che-motherapy. Our findings show that SSA can eliminateNSCLC cells through the induction of autophagy. Wepropose that this is a desirable property especially in thecase of lung cancer chemotherapy because drugs com-monly used in the clinic such as paclitaxel, gemcitabine,and EGF receptor (EGFR) inhibitors often lead to thedevelopment of apoptosis-resistant tumors (59). Recentstudies have shown that apoptosis resistance can developas a result of Akt or Bcl-2 upregulation and that thesetumors can be chemosensitized by inducers of autophagy(60–62).In summary, our findings support the potential of SSA
as a novel agent for the prevention and/or treatment oflung cancer. We provide proof-of-concept evidence thatthe induction of autophagy through theAkt/mTORpath-way may represent a largely unexplored antineoplasticproperty that could be targeted to develop safer andmoreefficacious NSAID derivatives. Further studies are there-fore warranted to determine the exact mechanism ofaction of SSA and to evaluate in vivo antitumor efficacyin suitable models of lung cancer. Identifying the specificmolecular target(s) has implications not only for thedevelopment of SSA and analogs but also can be used to
identify additional small molecules to develop moreselective and potent clinical candidates.
Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.
Authors' ContributionsConception and design: E. Gurpinar, W.E. Grizzle, J.J. Shacka, N.A.Piazza, A.B. Keeton, G.A. PiazzaDevelopment of methodology: E. Gurpinar, W.E. Grizzle, B.J. Mader, N.A. Piazza, A.B. KeetonAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): E. Gurpinar, W.E. Grizzle, A.B. KeetonAnalysis and interpretation of data (e.g., statistical analysis, biostatis-tics, computational analysis): E. Gurpinar, W.E. Grizzle, A.B. KeetonWriting, review, and/or revision of the manuscript: E. Gurpinar, W.E.Grizzle, J.J. Shacka, S. Russo, A.B. Keeton, G.A. PiazzaAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): E. Gurpinar, W.E. Grizzle, N. LiStudy supervision: W.E. Grizzle, J.J. Shacka, G.A. Piazza
AcknowledgmentsThe authors thank Drs. Robert Reynolds, Bini Mathew, and Kocharani
Jacob from Southern Research for the synthesis of sulindac sulfide amideandUAB FlowCytometry andHigh Resolution Imaging Core facilities fortheir excellent technical assistance.
Grant SupportThis research was supported by NIH grants NCI 1R01CA131378 and
1R01CA148817-01A1 (G.A. Piazza).The costs of publication of this article were defrayed in part by the
payment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.
Received July 30, 2012; revised January 28, 2013; accepted February 17,2013; published OnlineFirst February 26, 2013.
References1. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2012. CA Cancer
J Clin 2012;62:10–29.2. NCI. SEER Stat Fact Sheets: lung and bronchus. 2012. Available from:
http://seer.cancer.gov/statfacts/html/lungb.html3. NCI.National LungScreening Trial (NLST). 2012. Available from: http://
Miettinen OS. Survival of patients with stage I lung cancer detected onCT screening. N Engl J Med 2006;355:1763–71.
5. Apte RN, Krelin Y, Song X, Dotan S, Recih E, Elkabets M, et al. Effectsof micro-environment- and malignant cell-derived interleukin-1 incarcinogenesis, tumour invasiveness and tumour-host interactions.Eur J Cancer 2006;42:751–9.
6. Rothwell PM, Fowkes FG, Belch JF, OgawaH,WarlowCP,Meade TW.Effect of daily aspirin on long-term risk of death due to cancer: analysisof individual patient data from randomised trials. Lancet 2011;377:31–41.
7. Rothwell PM, Price JF, Fowkes FG, Zanchetti A, Roncaglioni MC,Tognoni G, et al. Short-term effects of daily aspirin on cancer inci-dence, mortality, and non-vascular death: analysis of the time courseof risks and benefits in 51 randomised controlled trials. Lancet 2012;379:1602–12.
8. Bowers WH, Spencer JB, McDevitt NB. Brachial-artery rupture inEhlers-Danlos syndrome: an unusual cause of high median-nervepalsy. A case report. J Bone Joint Surg Am 1976;58:1025–6.
9. Rothwell PM,WilsonM, Price JF, Belch JF,Meade TW,Mehta Z. Effectof daily aspirin on risk of cancermetastasis: a study of incident cancersduring randomised controlled trials. Lancet 2012;379:1591–601.
10. Khuder SA, Herial NA, Mutgi AB, Federman DJ. Nonsteroidal antiin-flammatory drug use and lung cancer: a metaanalysis. Chest 2005;127:748–54.
11. Rioux N, Castonguay A. Prevention of NNK-induced lung tumorigen-esis in A/J mice by acetylsalicylic acid and NS-398. Cancer Res1998;58:5354–60.
12. Duperron C, Castonguay A. Chemopreventive efficacies of aspirin andsulindac against lung tumorigenesis in A/J mice. Carcinogenesis1997;18:1001–6.
13. Malkinson AM, Koski KM, Dwyer-Nield LD, Rice PL, Rioux N, Cas-tonguay A, et al. Inhibition of 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone-induced mouse lung tumor formation by FGN-1 (sulindacsulfone). Carcinogenesis 1998;19:1353–6.
14. Vane JR, Bakhle YS, Botting RM. Cyclooxygenases 1 and 2. Annu RevPharmacol Toxicol 1998;38:97–120.
15. Tamirisa KP, Mukherjee D. Gene therapy in cardiovascular diseases.Curr Gene Ther 2002;2:427–35.
16. Kashfi K, Rigas B. Non-COX-2 targets and cancer: expanding themolecular target repertoire of chemoprevention. Biochem Pharmacol2005;70:969–86.
17. Whitt JD, LiN, TinsleyHN,ChenX, ZhangW, Li Y, et al. Anovel sulindacderivative that potently suppresses colon tumor cell growth by inhibit-ing cGMP phosphodiesterase and beta-catenin transcriptionalactivity. Cancer Prev Res (Phila) 2012;5:822–33.
18. Soh JW, Weinstein IB. Role of COX-independent targets of NSAIDsand related compounds in cancer prevention and treatment. Prog ExpTumor Res 2003;37:261–85.
19. He TC, Chan TA, Vogelstein B, Kinzler KW. PPARdelta is an APC-regulated target of nonsteroidal anti-inflammatory drugs.Cell 1999;99:335–45.
20. Piazza GA, Alberts DS, Hixson LJ, Paranka NS, Li H, Finn T, et al.Sulindac sulfone inhibits azoxymethane-induced colon carcinogene-sis in rats without reducing prostaglandin levels. Cancer Res 1997;57:2909–15.
Novel Sulindac Analog Induces Autophagy in Lung Cancer Cells
www.aacrjournals.org Mol Cancer Ther; 12(5) May 2013 673
21. Thompson HJ, Jiang C, Lu J, Mehta RG, Piazza GA, Paranka NS, et al.Sulfone metabolite of sulindac inhibits mammary carcinogenesis.Cancer Res 1997;57:267–71.
22. PiazzaGA, ThompsonWJ, PamukcuR,AlilaHW,WhiteheadCM, Liu L,et al. Exisulind, a novel proapoptotic drug, inhibits rat urinary bladdertumorigenesis. Cancer Res 2001;61:3961–8.
23. Chan DC, Earle KA, Zhao TL, Helfrich B, Zeng C, Baron A, et al.Exisulind in combinationwith docetaxel inhibits growth andmetastasisof human lung cancer and prolongs survival in athymic nude rats withorthotopic lung tumors. Clin Cancer Res 2002;8:904–12.
24. Whitehead CM, Earle KA, Fetter J, Xu S, Hartman T, Chan DC, et al.Exisulind-induced apoptosis in a non-small cell lung cancer orthotopiclung tumor model augments docetaxel treatment and contributes toincreased survival. Mol Cancer Ther 2003;2:479–88.
25. Berman KS, Verma UN, Harburg G, Minna JD, Cobb MH, Gaynor RB.Sulindac enhances tumor necrosis factor-alpha-mediated apoptosisof lung cancer cell lines by inhibition of nuclear factor-kappaB. ClinCancer Res 2002;8:354–60.
26. PiazzaGA, Keeton AB, Tinsley HN,Gary BD,Whitt JD,MathewB, et al.A novel sulindac derivative that does not inhibit cyclooxygenases butpotently inhibits colon tumor cell growth and induces apoptosis withantitumor activity. Cancer Prev Res (Phila) 2009;2:572–80.
27. Tinsley HN, Gary BD, Keeton AB, Zhang W, Abadi AH, Reynolds RC,et al. Sulindac sulfideselectively inhibits growth and induces apoptosisof human breast tumor cells by phosphodiesterase 5 inhibition, ele-vation of cyclic GMP, and activation of protein kinase G. Mol CancerTher 2009;8:3331–40.
28. Shacka JJ, Klocke BJ, Shibata M, Uchiyama Y, Datta G, Schmidt RE,et al. Bafilomycin A1 inhibits chloroquine-induced death of cerebellargranule neurons. Mol Pharmacol 2006;69:1125–36.
29. Klionsky DJ, Abeliovich H, Agostinis P, Agrawal DK, Aliev G, AskewDS, et al. Guidelines for the use and interpretation of assays formonitoring autophagy in higher eukaryotes. Autophagy 2008;4:151–75.
30. Ramser B, Kokot A, Metze D, Weiss N, Luger TA, Bohm M. Hydroxy-chloroquine modulates metabolic activity and proliferation andinduces autophagic cell death of human dermal fibroblasts. J InvestDermatol 2009;129:2419–26.
31. Gozuacik D, Kimchi A. Autophagy as a cell death and tumor suppres-sor mechanism. Oncogene 2004;23:2891–906.
32. Denton D, Nicolson S, Kumar S. Cell death by autophagy: facts andapparent artefacts. Cell Death Differ 2012;19:87–95.
33. Tsvetkov AS,Miller J, ArrasateM,Wong JS, PleissMA, Finkbeiner S. Asmall-molecule scaffold induces autophagy in primary neurons andprotects against toxicity in a Huntington disease model. Proc NatlAcad Sci U S A 2010;107:16982–7.
34. Mizushima N, Yoshimori T. How to interpret LC3 immunoblotting.Autophagy 2007;3:542–5.
35. Kimura S, Noda T, Yoshimori T. Dissection of the autophagosomematuration process by a novel reporter protein, tandem fluorescent-tagged LC3. Autophagy 2007;3:452–60.
36. Kimura S, Fujita N, Noda T, Yoshimori T. Monitoring autophagy inmammalian cultured cells through the dynamics of LC3. MethodsEnzymol 2009;452:1–12.
adaptations. Nat Cell Biol 2007;9:1102–9.39. Zhou H, Liu W, Su Y, Wei Z, Liu J, Kolluri SK, et al. NSAID sulindac and
its analog bind RXRalpha and inhibit RXRalpha-dependent AKT sig-naling. Cancer Cell 2010;17:560–73.
40. LeeHC, Park IC, ParkMJ, An S,WooSH, Jin HO, et al. Sulindac and itsmetabolites inhibit invasion of glioblastoma cells via down-regulationof Akt/PKB and MMP-2. J Cell Biochem 2005;94:597–610.
41. Hsu AL, Ching TT, Wang DS, Song X, Rangnekar VM, Chen CS. Thecyclooxygenase-2 inhibitor celecoxib induces apoptosis by blockingAkt activation in human prostate cancer cells independently of Bcl-2.J Biol Chem 2000;275:11397–403.
42. Zhu J, Huang JW, Tseng PH, YangYT, Fowble J, Shiau CW, et al. Fromthe cyclooxygenase-2 inhibitor celecoxib to a novel class of 3-phos-
phoinositide-dependent protein kinase-1 inhibitors. Cancer Res 2004;64:4309–18.
43. Turcotte S, Giaccia AJ. Targeting cancer cells through autophagy foranticancer therapy. Curr Opin Cell Biol 2010;22:246–51.
44. Saitoh M, Pullen N, Brennan P, Cantrell D, Dennis PB, Thomas G.Regulation of an activated S6 kinase 1 variant reveals a novel mam-malian target of rapamycin phosphorylation site. J Biol Chem 2002;277:20104–12.
45. Ramaswamy S, Nakamura N, Vazquez F, Batt DB, Perera S, RobertsTM, et al. Regulation of G1 progression by the PTEN tumor suppressorprotein is linked to inhibition of the phosphatidylinositol 3-kinase/Aktpathway. Proc Natl Acad Sci U S A 1999;96:2110–5.
46. Hoyer-Hansen M, Bastholm L, Szyniarowski P, Campanella M, Sza-badkai G, Farkas T, et al. Control of macroautophagy by calcium,calmodulin-dependent kinase kinase-beta, and Bcl-2. Mol Cell 2007;25:193–205.
47. Grossman SR, Perez M, Kung AL, JosephM, Mansur C, Xiao ZX, et al.p300/MDM2 complexes participate in MDM2-mediated p53 degrada-tion. Mol Cell 1998;2:405–15.
48. Chiou SK, Hoa N, Hodges A. Sulindac sulfide induces autophagicdeath in gastric epithelial cells via survivin down-regulation: a mech-anism of NSAIDs-induced gastric injury. Biochem Pharmacol 2011;81:1317–23.
49. RazA. Is inhibition of cyclooxygenase required for the anti-tumorigeniceffects of nonsteroidal, anti-inflammatory drugs (NSAIDs)? In vitroversus in vivo results and the relevance for the prevention and treat-ment of cancer. Biochem Pharmacol 2002;63:343–7.
50. Williams CS, Watson AJ, Sheng H, Helou R, Shao J, DuBois RN.Celecoxib prevents tumor growth in vivowithout toxicity to normal gut:lack of correlation between in vitro and in vivo models. Cancer Res2000;60:6045–51.
51. Piazza GA, Rahm AK, Finn TS, Fryer BH, Li H, Stoumen AL, et al.Apoptosis primarily accounts for the growth-inhibitory properties ofsulindacmetabolites and involves amechanism that is independent ofcyclooxygenase inhibition, cell cycle arrest, and p53 induction. CancerRes 1997;57:2452–9.
52. Piazza GA, Rahm AL, Krutzsch M, Sperl G, Paranka NS, Gross PH,et al. Antineoplastic drugs sulindac sulfide and sulfone inhibit cellgrowth by inducing apoptosis. Cancer Res 1995;55:3110–6.
53. ZimmermannKC,WaterhouseNJ,Goldstein JC,SchulerM,GreenDR.Aspirin induces apoptosis through release of cytochrome c frommitochondria. Neoplasia 2000;2:505–13.
54. Levine B, Kroemer G. Autophagy in the pathogenesis of disease. Cell2008;132:27–42.
55. Wu WK, Coffelt SB, Cho CH, Wang XJ, Lee CW, Chan FK, et al. Theautophagic paradox in cancer therapy. Oncogene 2012;31:939–53.
56. Kroemer G, Levine B. Autophagic cell death: the story of a misnomer.Nat Rev Mol Cell Biol 2008;9:1004–10.
57. SalazarM,CarracedoA,Salanueva IJ,Hernandez-TiedraS, LorenteM,Egia A, et al. Cannabinoid action induces autophagy-mediated celldeath through stimulation of ER stress in human glioma cells. J ClinInvest 2009;119:1359–72.
58. Shimizu S, Kanaseki T, Mizushima N,Mizuta T, Arakawa-Kobayashi S,Thompson CB, et al. Role of Bcl-2 family proteins in a non-apoptoticprogrammed cell death dependent on autophagy genes. Nat Cell Biol2004;6:1221–8.
59. d'Amato TA, Landreneau RJ, Ricketts W, HuangW, Parker R, Mechet-ner E, et al. Chemotherapy resistance and oncogene expression innon-small cell lung cancer. J Thorac Cardiovasc Surg 2007;133:352–63.
60. Barbone D, Ryan JA, Kolhatkar N, Chacko AD, Jablons DM, Sugar-baker DJ, et al. The Bcl-2 repertoire ofmesothelioma spheroids under-lies acquired apoptotic multicellular resistance. Cell Death Dis 2011;2:e174.
61. Powell CA. Rounding up apoptosis resistance targets in lung cancer.Am J Respir Cell Mol Biol 2009;41:7–8.
62. Kim KW, Hwang M, Moretti L, Jaboin JJ, Cha YI, Lu B. Autophagyupregulation by inhibitors of caspase-3 and mTOR enhances radio-therapy in a mouse model of lung cancer. Autophagy. 2008;4:659–68.
Gurpinar et al.
Mol Cancer Ther; 12(5) May 2013 Molecular Cancer Therapeutics674
2013;12:663-674. Published OnlineFirst February 26, 2013.Mol Cancer Ther Evrim Gurpinar, William E. Grizzle, John J. Shacka, et al. of AutophagyGrowth through Suppression of Akt/mTOR Signaling and Induction A Novel Sulindac Derivative Inhibits Lung Adenocarcinoma Cell
Updated version
10.1158/1535-7163.MCT-12-0785doi:
Access the most recent version of this article at:
