8-amino-adenosineactivatesp53-independentcelldeathof metastatic breast … · breast cancer cell...

Preclinical Development

8-Amino-AdenosineActivates p53-IndependentCell Death ofMetastatic Breast Cancers

Alla Polotskaia1, Sandy Hoffman1, Nancy L. Krett2, Mala Shanmugam2, Steven T. Rosen2, and Jill Bargonetti1

Abstract8-Amino-adenosine (8-NH2-Ado) is a ribose sugar nucleoside analogue that reduces cellular ATP levels and

inhibits mRNA synthesis. Estrogen receptor-negative (ER�) metastatic breast cancers often contain mutant

p53; therefore, we asked if 8-NH2-Ado could kill breast cancer cells without activating the p53-pathway.

Regardless of the breast cancer subtype tested or the p53 status of the cells, 8-NH2-Ado was more cytotoxic

than either gemcitabine or etoposide. 8-NH2-Ado treatment inhibited cell proliferation, activated cell death,

and did not activate transcription of the p53 target gene p21 or increase protein levels of either p53 or p21. This

occurred in the estrogen receptor-positive (ERþ) MCF-7 cells that express wild-type p53, the ERþ T47-D cells

that express mutant p53, and the ER�MDA-MB-468 cells orMDA-MB-231 cells that both express mutant p53.

8-NH2-Ado induced apoptotic death of MCF-7 cells and apoptosis was not inhibited by knockdown of

functional p53. Moreover, the pan-caspase inhibitor Z-VAD blocked the 8-NH2-Ado–induced MCF-7 cell

death. Interestingly, 8-NH2-Ado caused the MDA-MB-231 cells to detach from the plate with only limited

evidence of apoptotic cell death markers and the cell death was not inhibited by Z-VAD. Inhibition of MDA-

MB-231 cell autophagy, by reduction of ATG7 or 3-methyladenine treatment, did not block this 8-NH2-Ado–

mediated cytotoxicity. Importantly 8-NH2-Ado was highly cytotoxic to triple-negative breast cancer cells and

worked through a pathway that did not require wild-type p53 for cytoxicity. Therefore, 8-NH2-Ado should be

considered for the treatment of triple-negative breast cancers that are chemotherapy resistant.Mol Cancer Ther;

11(11); 2495–504. �2012 AACR.

IntroductionTreatment of metastatic and triple-negative breast can-

cer remains a challenge in the clinical setting.Gemcitabineis a nucleoside analog that is approved by the U.S. Foodand Drug Administration to be used in combination withpaclitaxel for the first-line treatment of patients withmetastatic breast cancer after they have received anthra-cycline chemotherapy. Gemcitabine activates DNA dam-age pathways that signal to the tumor suppressor p53 (1),and a functional p53 pathway improves gemcitabinecytotoxicity (2). The p53 protein is a tumor suppressorthat activates transcription of many downstream targetgenes and thereby controls cell growth and cell death (3).However, in 70% of human cancers, the p53 gene is foundmutated (4). In breast cancer, p53 is often foundmutated in

triple-negative andmetastatic tumors (5). Therefore, iden-tifying nucleoside analog chemotherapeutic treatmentsthat can kill breast cancers independently of activating thep53 pathway is an important aspect of improving cancertreatments. The identification of drugs that can inhibit thegrowth of triple-negative breast cancers with mutant p53is an understudied area. Therefore, we have worked toidentify a compound that inhibits breast cancer prolifer-ation regardless of the hormone receptor status or themutant p53 status. We hypothesized that the nucleosideanalogue, 8-amino-adenosine (8-NH2-Ado)would inhibitbreast cancer cell proliferation independently of the p53pathway.

In normal cells, wild-type p53 is the guardian of thegenome, however, inactivation of p53 through eithermutation or interactions with oncogenic proteins, playsa large role in tumor promotion (4). Currently, breastcancers resistant to chemotherapy are treated withdrugs that activate DNA damage pathways (6). DNAdamaging drugs activate the p53 pathway and studiesindicate that wild-type p53 can be responsible for thevast number of side effects associated with the death ofnormal cells during DNA damage (7). Recent work witha panel of breast cancer cell lines suggests that combin-ing the DNA-damaging agent gemcitabine with a PARPinhibitor (that blocks DNA repair and inhibits all cellcheckpoints) enhances the cytotoxic outcome on cancercells (8–10). However, a chemotherapeutic nucleoside

Authors' Affiliations: 1The Department of Biological Sciences HunterCollege and The Graduate Center Departments of Biology and Biochem-istry, CUNY, New York, New York; and 2Division of Hematology Oncology,Department of Medicine, Robert H. Lurie Comprehensive Cancer Center,Feinberg School of Medicine, Northwestern University Medical School,Chicago, Illinois

Corresponding Author: Jill Bargonetti, Department of BiologicalSciences, Hunter College, CUNY, 695 Park Ave., New York, NY 10065.Phone: 212-650-3519; Fax: 212-772-5227; E-mail:[email protected]

doi: 10.1158/1535-7163.MCT-12-0085

�2012 American Association for Cancer Research.

MolecularCancer

Therapeutics

www.aacrjournals.org 2495

that does not activate the p53 pathway might be moreeffective.

Unique and promising RNA-directed treatments formyeloma are the nucleoside analogs 8-chloro-adenosine(currently in phase I trials) and 8-NH2-Ado (11, 12).Importantly, 8-NH2-Ado is more cytotoxic to myelomacancer cells than to normal human lymphocytes (13, 14).Moreover, 8-NH2-Ado is an RNA-directed nucleosideanalog that inhibits transcription and polyadenylation(12, 15). Adenosine analogues accumulate in cells astriphosphates and decrease the endogenous ATP pool(12). Preclinical studies on multiple myelomas havebeen very encouraging (13, 14, 16, 17). 8-Chloro-aden-osine has recently proved cytotoxic to breast cancer cellsthrough depletion of cyclin E (11). However, no studyhas investigated whether 8-NH2-Ado can effectively killbreast cancer cells without activating a p53 pathway.We have investigated whether p53 function is requiredfor 8-NH2-Ado to inhibit breast cancer cell proliferation.We predicted that 8-NH2-Ado would inhibit the growthof breast cancer cells without requiring functional p53.In this report, we show that 8-NH2-Ado is more effec-tive than gemcitabine and etoposide at inhibiting thegrowth of multiple breast cancer subtypes, includingtriple-negative cell lines. Furthermore, we show that 8-NH2-Ado induces p53-independent cell death that canproceed through an apoptotic pathway as well througha novel cytotoxic pathway that does not require autop-hagy or necrosis.

Materials and MethodsCell culture

All the cell lines used for this study were obtainedfrom American Type Culture Collection (ATCC). Theauthors did not carry out further authentication. MDA-MB-231 cells (p53 mutant 280, Arg–Lys), T47D (p53mutant 194, Leu–Phe), MDA-MB-468 (p53 mutant 273,Arg–His), and MCF-7 (p53 wild-type) from ATCC weregrown in Dulbecco’s Modified Eagle Medium (DMEM)medium (Mediatech), 10% FBS (Gemini), and 50 U/mLof penicillin–streptomycin (Mediatech) at 5% CO2 in a37�C humidified incubator. Clonal MCF-7 line D11 withinducible (Tet-on) shRNA for p53 was established andcharacterized in our laboratory (18). To induce shRNAexpression, D11 cells were treated with 2 mg/mL doxy-cycline for 6 days. The MCF7.beclin1 clone (19) was agenerous gift from Beth Levine (UT Southwestern Med-ical Center, Dallas, TX) and was grown in the absence oftetracycline to induce Beclin 1 expression for 5 days.

ReagentsEtoposide, propidium iodide, doxycyclin, trypan blue

solution, and MTT)were purchased from Sigma. Gemci-tabine and 8-NH2-Adowere provided by Dr. Steve Rosen(Robert H. Lurie Comprehensive Cancer Center, Chicago,IL). The activity of the 8-NH2-Ado varied in the 2 batchesdonated to the laboratory. Therefore, the treatments toachieve similar inhibition of growth ranged from 10 to 15

mmol/L. The general Caspase Inhibitor Z-VAD-FMKwaspurchased from R&D Systems (FMK001). The Image-iTLIVE Lysosomal and Nuclear Labeling Kit (MolecularProbes I34202) was used to detect autophagy in the cells.Cell Death Detection ELISA was purchased from Roche(11 544 675 001). Apoptosis detection was done withApoScreen Annexin V-FITC Kit (Southern Biotech10010-02). 3-Methyladenine (3-MA) and Necrostatin 1were purchased from Sigma.

MTT analysis, flow cytometry, and apoptosisdetection by Histone: DNA complexes

MTT. Toxic effects of the drugs were determined bymeasuring the mitochondrial activity of each cell lineusing the tetrazolium dye-based microtitration assay tomeasure mitochondrial dehydrogenases activity (20).Cells were seeded at 1.25 � 105 cells per well in 12-wellplates and allowed to attach overnight. Cells were thentreated with the drugs as indicated in the figures at theconcentrations shown. The absorbance was quantified bymeasuring the absorbance at 550 nm (the 620 nm absor-bance was subtracted for background). All MTT data arerepresented as mitochondrial dehydrogenase activity aspercentage of a dimethyl sulfoxide (DMSO) vehicle-trea-ted control.

Flow cytometry. The Annexin V-FITC reagent wasused according to manufacturer’s protocol after 16 hoursof drug treatment. Cells were detached with trypsin,washed twice with PBS, and resuspended in the bindingbuffer provided by the manufacturer with the addition of10 mL ApoScreen Annexin V-FITC and incubated for 10minutes. This stepwas followedby the additionof 10mLofpropidium iodide and then flow cytometry was con-ducted. Fluorescence-activated cell sorting (FACS) anal-ysis was carried out using on a BD Bioscience FACS scan.

Enrichment factor detection. The apoptosis detectionkit used anti-histone antibody and anti-DNA antibody todetect the values of histones associated with DNA in thecytoplasm as an indication of apoptosis. Cytoplasmicextract was prepared after the cells were treated withdrugs for 24 hours in a 12-well plate. The plate was spundown at 1,500 � g for 10 minutes and the cells wereresuspended in 500 mL of lysis buffer and incubated for30 minutes at room temperature. The cells were resus-pended and spun in an eppendorf tube at 13,000 rpm for10minutes and the supernatantwas collected andused forthe ELISA procedure at a 1:3 dilution. The ELISA stepswere carried out as described by the manufacturer’sdirection with slight modifications. The final step wasmodified to stop color development by the addition of 100mL of 5% SDS. Detectionwas carried out at 415 nm againsta substrate solution blank. Enrichment factor¼mUof thedrug treated (dying/dead cells)/mU of the correspond-ing control.

Quantitative reverse transcription PCRStandardprocedures fromthemanufacturerswereused

for these assays. RNA was isolated using QIAshredder

Polotskaia et al.

Mol Cancer Ther; 11(11) November 2012 Molecular Cancer Therapeutics2496

columns and RNeasy Mini Kit (Qiagen). Five microgramsof RNAwas used for cDNA synthesis using high capacitycDNA Archive Kit reagents (Applied Biosystems). Onehundred and fifty nanograms of cDNA was combinedwith TaqMan Universal Master Mix and Applied Biosys-tems Assays on Demand primers/probes for p21(Hs00355782_m1) or actin (4352935E). PCR reaction wascarried out in 7500 Sequence Detection System (AppliedBiosystems) and actin was used as the normalizer.

Whole-cell protein extractionCells were lysed in radioimmunoprecipitation assay

(RIPA) buffer (0.1% SDS, 1% NP-40, 0.5% deoxycholate,150 mmol/L NaCl, 1 mmol/L EDTA, 0.5 mmol/LEGTA, 50 mmol/L Tris-Cl pH8) with 1 mmol/L phe-nylmethylsulfonylfluoride (PMSF), 8.5 mg/mL aproti-nin, and 2 mg/mL leupeptin and incubated on ice for 20minutes with periodic vortexing. Pellets were then cen-trifuged at 9,300 � g for 20 minutes. The supernatantswere collected and kept at �80�C for future analysis.

Western blot analysisFifty micrograms of protein extract were separated by

4% to 12% SDS-PAGE (Invitrogen) and electrotransferredto nitrocellulose or polyvinylidene difluoride (PVDF)membrane. Immunoblotting was done with p53 mono-clonal antibody supernatants (pAb421, pAb240, andpAb1801), LC3B (Cell Signaling; 2775), PARP (BD Biosci-ence; 51-6639GR), p21 (Ab-1 Oncogene Research Science;OP64), Beclin 1 (Novus Biologicals; 110-87318), ATG-7(Cell Signaling; 8558), and actin (Sigma; A2066). Themembranes were then incubated in anti-mouse or anti-rabbit horseradish peroxidase–conjugated secondaryantibodies (Sigma) and the signals were visualized bychemiluminescence.

Lysosomal stainingLabeling of the live cells adherent to the cover slipswith

red-fluorescent LysoTracker dye and blue-fluorescentHoechst dye was conducted according to manufacturer’sprotocol and images were taken on an Olympus fluores-cent microscope.

RNA interference and transfectionsFor siRNA experiments, MDA-MB-231 cells were seed-

ed inmediawithout penicillin–streptomycin and allowedto attach overnight. Cells were transfected at 60% con-fluency with 100 nmol/L of atg7 siRNA smart pool ornontargeted siRNA from Dharmacon for 6 hours usingLipofectamine 2000 (Invitrogen) as per manufacturersprotocol.At the endof the incubationperiod, freshDMEMwith 10% FBS was added and the cells were allowed togrow for 48hours.Drug treatmentwith 15mmol/L8-NH2-Ado in freshDMEMwith 10%FBSwas then carriedout for24 hours. Cellswere harvested by scraping into themedia,washed with PBS, and lysed in RIPA buffer for proteinanalysis or analyzed on the plate by MTT analysis forproliferation.

Results8-NH2-Ado is more cytotoxic than standardchemotherapeutic drugs to breast cancer cell linesregardless of p53 or estrogen receptor status

Cell culture studies are an excellent first-line indicationof the efficacy of chemotherapeutic drugs for particularcancer subtypes. Therefore, cell culture studies with well-defined cell lineswith defined genotypes are ideal for firstline preclinical studies. We used a well-defined set ofbreast cancer cell lines ranging from tumorigenic to highlymetastatic and compared their sensitivity with 8-NH2-Ado and with the 2 DNA-damaging drugs (gemcitabineand etoposide). The cell lines examined were MCF-7,MDA-MB-231, T47-D, and MDA-MB-468. These 4 celllines are well described for their p53 mutations (5).MCF-7 cells have wild-type p53, whereas MDA-MB-231, T47-D, and MDA-MB-468 cells have mutant p53. Inaddition, MCF-7 and T47-D cell lines are both estrogenand progesterone receptor positive. We compared theoutcome of cells treated with 8-NH2-Ado with thatobserved with the standard chemotherapeutics gemcita-bine andetoposide.Gemcitabine is aDNA-directednucle-oside analog that causes replication stalling and single-stranded breaks in the DNA, whereas etoposide is atopoisomerase inhibitor that causes double-strandedbreaks to the DNA (1). In contrast, 8-NH2-Ado is anRNA-directed nucleoside analog (12). Treatments werecarried out with relevant concentrations for gemcitabine(8) and etoposide (18) in breast cancer and for 8-NH2-Adoin myeloma (13). Treatment with 10 mmol/L 8-NH2-Adoapproached the IC50 dosage for MCF-7 cells and sur-passed the IC50 value for the 3 cell lines with mutantp53 (MDA-MB-231, T47-D, and MDA-MB-468 cells; Fig.1). TheMTTproliferation assaymonitors cell populations’response to external factors including cell growth and celldeath. In all instances with 8-NH2-Ado, the cells detachedfrom the plate (most likely indicating cell death, see Fig.4A for representative images). Moreover, in all cases, 8-NH2-Ado treatment was far more cytotoxic than eitheretoposide or gemcitabine (Fig. 1). This strongly suggestedthat functional p53 was not required for cell death induc-tion. To examine if themechanism of action of 8-NH2-Adowas p53-independent, we focused on comparing theMCF-7 cell line with the MDA-MB-231 cell line. These 2cell lines were uniquely sensitive to 8-NH2-Ado. MDA-MB-231 cells are metastatic and express gain-of-functiononcogenic mutant p53 that blocks the p53-family memberprotein p63 (21), whereas MCF-7 cells have wild-type p53and are not metastatic (22).

In contrast to standard chemotherapeutics,8-NH2-Ado does not activate the p53 pathway

Wild-type p53 is normally found at low levels in divid-ing cells due to the targeted destruction by its negativeregulator Mdm2 (23). After DNA damage, the levels ofwild-type p53 protein increase due to posttranslationalmodifications that block the interaction of p53withMdm2(24). High levels of oncogenic variant p53 protein are a

8-Amino-Adenosine Induces p53-Independent Cell Death

www.aacrjournals.org Mol Cancer Ther; 11(11) November 2012 2497

hallmark of cancer cells containing genetic point muta-tions in the p53 gene that results in stable p53 protein (25).We detected an increase inwild-type p53 protein inMCF-7 cells after etoposide and gemcitabine treatment (Fig. 2A,lanes 2 and 4); however, nowild-type p53 protein increasewas detected after 8-NH2-Ado treatment (Fig. 2A, com-pare lanes 1 and 3). Furthermore, MDA-MB-231 cellscontained high levels of stable mutant p53 before drugtreatment and this protein level was not affected byetoposide, 8-NH2-Ado, or gemcitabine treatment (Fig.2A, lanes 5–8).

To further examine the p53-independent signaling of8-NH2-Ado, we asked if drug treatment could increasethe level of the cyclin-dependent kinase inhibitor p21.We reproducibly observed that 8-NH2-Ado treatmentreduced the level of p21 protein in MCF-7 cells anddecreased the p21 transcript level in both MCF-7 andMDA-MB-231 cells (Fig. 2). This indicated that 8-NH2-Ado did not activate a DNA damage response in eithercell line and blocked the transcription of a key cell-cycleinhibitor. This was in stark contrast to the outcomesobserved with etoposide or gemcitabine treatment, bothof which activated the transcription of p21. As expected,etoposide mediated a robust increase in p21 protein andtranscript in MCF-7 cells and surprisingly increased p21protein and transcript levels in MDA-MB-231 cells (Fig.2). Gemcitabine treatment of both cell lines caused a

robust increase in p21 transcript without a significantchange in p21 protein (Fig. 2). In support of our p53-independent hypothesis, no signaling to p53 was evi-dent in 8-NH2-Ado–treated cells, whereas both etopo-side and gemcitabine caused DNA damage signalingthat could function through wild-type p53 in MCF-7cells as well as through an alternative pathway in theMDA-MB-231 cells, perhaps through the p53 familymember p73.

8-NH2-Ado induces significant apoptotic death ofMCF-7 cells

To determine the signal transduction pathways activat-ed after treatment of the cells with 8-NH2-Ado, weexamined cell death markers associated with apoptosis.We assessed the cleavage of PARP, the enrichmentof cytoplasmic histone-associated DNA fragments, andAnnexinV staining (Fig. 3).WedetectedPARPcleavage inMCF-7 cells treated with 8-NH2-Ado (Fig. 3A, comparelane 3 with lanes 1, 2, and 4), as well as an increasein apoptosis-associated enrichment factor, which detectscytoplasmic histones attached to fragmented DNA (Fig.3B) and a reproducibly robust increase in Annexin Vstaining (Fig. 3C). In MDA-MB-231 cells, we observed areduction in PARP (most likely through degradation)and found no detectable cleavage product after treatmentwith 8-NH2-Ado (Fig. 3A, compare lane 6 with lanes 4, 5,

Figure 1. 8-NH2-Ado is morecytotoxic than the standardchemotherapeutic drugs to breastcancer cell lines with either wild-type or mutant p53. Breast cancercell lines with either wild-typep53 (A, MCF-7) or oncogenicmutant p53 (B, MDA-MB-231;C, T47-D; and D, MDA-MB-468)were incubated with etoposide(ETOP), 8-NH2-Ado (8AA), orgemcitabine (GEM). E, structuresof compounds used (obtained fromchemicalbook.com). All drugtreatments were carried out at aconcentration of 10 mmol/L for 24hours. Sensitivity of the cells to the3 drugs was assessed by MTTassay based on mitochondrialdehydrogenase activity. SE barsrepresent 3 independentexperiments. Results werenormalized to the DMSOtreatment.

Polotskaia et al.


and 7).Moreover, in theMDA-MB-231 cells, wedetected areproducibly low increase in apoptosis enrichment factorand Annexin V staining (Fig. 3B and C), suggesting thatthe death of these cells was through a nonapoptoticpathway.

The pan-caspase inhibitor Z-VADblocks 8-NH2-Adoinduced MCF-7 death, but not MDA-MB-231 celldeathTo further examine the apoptotic cell death of MCF-7

cells, we asked whether the observed cell death could beinhibited with the pan-caspase inhibitor Z-VAD-FMK(Z-VAD). The change in morphology of 8-NH2-Ado–treated MCF-7 cells (including the floating cell pheno-type) was completely blocked when Z-VAD was simul-taneously added to the treatment protocol (Fig. 4A).Moreover, simultaneous addition of Z-VAD blocked thecleavage of PARP (Fig. 4B, compare lanes 2 and 3).However, in MDA-MB-231 cells, the simultaneous addi-

tion of Z-VAD did not inhibit 8-NH2-Ado–inducedfloating cells or PARP reduction (Fig 4A and B). Ourdata indicate that the 8-NH2-Ado–induced death ofMCF-7 cells was caspase dependent, whereas 8-NH2-Ado induced death of MDA-MB-231 cells was caspaseindependent.

The p53 protein is not required for 8-NH2-Ado–induced MCF-7 cell death

To confirm that the killing ability of 8-NH2-Ado did notrequire functional p53, we testedwhether the knockdownof p53 by inducible shRNAwould influence the apoptosisoutcome ofMCF-7 cells. We have previously reported theconstruction of an inducible shRNA p53 knockdown cellline of MCF-7 cells (18). We used this MCF-7.shp53 2120clone (D11) for this study.Knockdownofp53was inducedby the addition of doxycycline (Fig. 5A). No detectablechange in 8-NH2-Ado–induced PARP cleavage orMCF-7.shp53 clone viability was caused by the reduction in p53protein (Fig. 5). In combinationwith the data showing that8-NH2-Ado did not activate the wild-type p53 pathway(Fig. 2), these results support the conclusion that 8-NH2-Ado–induced apoptotic cell death of MCF-7 cells is p53independent. However, this did not address the mecha-nism of cell death occurring in the MDA-MB-231 cells.Previous studies have detected that some novel chemo-therapeutic drugs induce autophagic cell death of MDA-MB-231 cells (26, 27). We addressed the possibility that 8-NH2-Ado increased autophagy in MDA-MB-231 cellsuntil the point of cell death.

8-NH2-Ado increases the autophagy markers inMDA-MB-231 cells but does not require autophagyfor cytotoxicity

Although autophagy is initially a cell survival pathwaywhere cytosolic components are encapsulated in double-membrane vesicles, it can also be used as a death pathwayif pushed to cannibalize the cell (28, 29). We examined theincrease of 2 different autophagic markers in the cellsbefore and after 8-NH2-Ado treatment. 8-NH2-Ado treat-ment reduces glucose consumption and myeloma cellscounteract this stress by increasing autophagy (16). It ispresumed that autophagy is a prosurvival response of themyeloma cells and not a cell death mechanism. Duringautophagy, cytoplasmic LC3-I ismodified to becomeLC3-II (30). The processing of LC3-I to LC3-II is visible as amore quickly migrating form on SDS-PAGE. However,the increase in LC3-II can be assessed as a marker forinduction of autophagy or inhibition of autophagosomeclearance (31, 32). We saw an increase in LC3-II in MDA-MB-231 cells treatedwith 8-NH2-Adoanda slight increasein LC3-II in the treated MCF-7 cells (Fig. 6A, comparelanes 1 and 2 with lanes 3 and 4). We also observed anincrease in acidic organelles in the treated MDA-MB-231cells using LysoTracker Red staining (data not shown).

Although the appearance of increased LC3-II indicatesthat during the cell death there is an associated autophagyresponse, it cannot be used to claim that 8-NH2-Ado

Figure2. In contrast to standard chemotherapeutics, 8-NH2-Adodoesnotincrease protein levels of either p53 or p21 or transcription of the p53target gene p21. MCF-7 and MDA-MB-231 cells were treated with 10mmol/L etoposide (ETOP), 8-NH2-Ado (8AA), or gemcitabine (GEM) for 24hours. A, Western blot analysis was used to assess the level of p53, p21,and actin protein from whole-cell lysates of MCF-7 and MDA-MB-231cells (as indicated). Quantitative real-time PCR was carried out todetermine the fold increase of p21mRNA after drug treatment relative tothe DMSO-treated control in MCF-7 cells (B) and MDA-MB-231 cells (C).Normalized to actin and DMSO vehicle cells.



induces autophagic cell death. We therefore asked howincreased expression of the autophagy inducing protein,Beclin-1, in MCF-7 cells influenced the 8-NH2-Ado–induced death pathway. We used MCF7.beclin1 clonesthat were a generous gift from Beth Levine (19). Theincrease in Beclin-1 expression in MCF-7 cells did notcause an increase in LC3-II (Fig. 6A, lanes 5 and 6).Importantly, MCF-7 cells have low expression of theautophagy regulating protein, Beclin 1 (19). The introduc-tion of exogenously expressed Beclin 1 into MCF-7 cellspromotes autophagy, inhibits cellular proliferation, andblocks the tumorigenesis of these cells in nude mice (19).We treated Beclin 1 overexpressing MCF-7 cells with8-NH2-Ado to determine whether this influenced theMCF-7 cell death. Although 8-NH2-Ado induced PARPcleavage in MCF-7 cells, we found that Beclin 1 over-expression inMCF7.beclin1 clones blocked the 8-NH2-AdoPARP cleavage (Fig. 6A, compare lanes 3 and 4with lanes5 and 6). TheMCF7.beclin1 cells were resistant to 8-amino-adenosine–mediated apoptosis as indicated by a lack offloating cells (data not shown) and no evident PARPcleavage (Fig. 6A, lanes 5 and 6). However, even thoughPARP cleavage was not apparent, 8-NH2-Ado–treatedMCF7.beclin1 cells still exhibited decreased proliferationand reduced viability (seen by MTT and trypan blueexclusion, data not shown). This suggested that autop-hagy accompanied the 8-NH2-Ado–induced cell deathbut did not assist the death or make it more aggressive.

To further examine the influence of autophagy on8-NH2-Ado–induced cell death, we observed treated

MDA-MB-231 cells during inhibition of autophagy bypharmacologic treatment using 3-MA or by siRNAatg7–mediated knockdown (Fig. 6 B–F). Simultaneousaddition of 8-NH2-Ado with 10 mmol/L of 3-MA did notinfluence the PARP Western blot pattern (Fig. 6B). Thetreatment with 3-MA slightly decreased the viability asassessed by theMTT assay and the appearance of floatingcells (Fig. 6C andD).We also knockeddown atg7 to inhibitautophagy and saw no increase in MTT activity (Fig. 6Eand F). The knockdown of atg7 increases MCF-7 cellsresistance to photodynamic therapy suggesting that ithelps the cells to die (33). However, 8-NH2-Ado treatmentcaused a p53-independent cell death that was associatedwith autophagy but was not assisted by autophagy. Thiscorresponds with the previous data showing 8-NH2-Adoinduces autophagy as a survival response (16). Further-more, we examined if the deathwas caused by necrosis bypretreating the MDA-MB-231 cells with a pharmacologicinhibitor called necrostatin 1. This also did not inhibit the8-NH2-Ado induced cell death (data not shown).

DiscussionApoptosis is a well-described cell death pathway but

the relationship between autophagy and cell death iscontroversial (34). The autophagy gene beclin 1 is a hap-loinsufficient tumor suppressor and increased expressionof Beclin 1 in MCF-7 cells promotes autophagy and inhi-bits the formation of human breast tumors in mousemodels (35). Autophagic cell death is sometimesdescribed as programmed cell death II, and has been

Figure 3. 8-NH2-Ado induces robust apoptotic cell death inMCF-7, but not inMDA-MB-231 cells. MCF-7 andMDA-MB-231 cells were treatedwith 10 mmol/Letoposide (ETOP), 8-NH2-Ado (8AA), or gemcitabine (GEM) for 24 hours. A, Western blot analysis was used to assess PARP cleavage and actin protein fromwhole-cell lysates of MCF-7 andMDA-MB-231 cells (as indicated). B, an ELISA for the quantitative, in vitro, determination of cytoplasmic histone-associatedDNA fragments was carried out. The values were scored as enrichment factor for the comparison of absorbance from the drug treatment of MCF-7andMBA-MB-231cells versus the control DMSOvehicle–treated cells. Increasedenrichment factor values in thehistogramserveasan indicationof increasedapoptosis. C, to assess for early-stage apoptosis, MCF-7 and MDA-MB-231 cells were treated with 10 mmol/L 8-NH2-Ado for 16 hours, scraped fromthe plate for harvesting, and stained with Annexin V and propidium iodide. The bottom right quadrant shows the early apoptotic cells that are AnnexinV positive. In MCF-7 cells (top), there were 4.1% early Annexin V–positive cells before treatment and 25.2% positive cells after treatment. In MDA-MB-231cells (bottom), there were 1.7% early Annexin V–positive cells before treatment and 6.4% after treatment.

Polotskaia et al.


suggested as a new target for cancer therapy (36). How-ever, it has been clearly documented that autophagymaintains cellular homeostatsis and can have both anti-tumor and tumor-promoting properties (37). Recently,autophagy has been shown to promote ras-driven tumorgrowth (38). 8-NH2-Ado has previously been associatedwith a survival program that is initiated because of met-

abolic dysfunction and this survival program happensalongside the induction of apoptosis (16). In this study,we investigated if p53 function played a part in 8-NH2-Ado–induced cell death. Importantly,wehavediscoveredthat 8-NH2-Ado inhibits the growth of multiple breastcancer lines without engaging a functional p53 program.Moreover, we have made the important observation that

Figure 4. The pan-caspase inhibitor Z-VAD blocks 8-NH2-Ado–induced PARP cleavage. MCF-7 and MDA-MB-231 cells were treated with 10 mmol/L8-NH2-Ado for 24 hours in the presence or absence of 50 mmol/L Z-VAD, the pan-caspase inhibitor added to the cell growth media. A, phasecontrast microscopy at �20 magnification showed a reversion of MCF-7 cells, but not MDA-MB-231 cells, to the DMSO vehicle–treated cell morphology(with fewer floating cells) when Z-VAD was added with 8-NH2-Ado. B, Western blot analysis of the protein extract from MCF-7 and MDA-MB-231 cellstreatedwith 8-NH2-Ado andZ-VADwas used to determine if Z-VAD addition reversed the PARPcleavage. Lanes are as indicated in the figure. C, the structureof Z-VAD was obtained from chemicalbook.com.

Figure 5. The p53 protein is notrequired for 8-NH2-Ado–inducedMCF-7 cell death. The MCF-7 p53shRNA 2120 clone D11 was treatedwith 2 mg/mL doxycycline (þDOX) for6 days of induced shRNA expressionandmaximum reductionof p53.Cellswith or without p53 knockdown weretreated with 10 mmol/L 8-NH2-Adofor 24 hours. A,Western blot analysiswas used to assess the level of p53,PARP, and actin protein from whole-cell lysates of the clonal p53 shRNAMCF-7 cells as indicated. B,sensitivity of the cells to 8-NH2-Adowith or without p53 knockdown wasassessed by MTT assay of live cellsbased on mitochondrialdehydrogenase activity and is shownas percentage of DMSO-treatedcontrol. C, the structure of DOX wasobtained from chemicalbook.com.



8-NH2-Ado can initiate a caspase-dependent cell deathpathway in MCF-7 cells and a caspase-independent celldeath pathway in MDA-MB-231 cells. This suggests that8-NH2-Ado is a possible therapeutic option for cancerswith mutant p53 as well as for certain cancers lackingfunctional apoptotic pathways.

Surprisingly, 8-NH2-Ado treatment of different breastcancer cell lines inhibited cancer cell growth by differentpathways. 8-NH2-Ado treatment was unable to activatewild-type p53. This is consistent with 8-NH2-Ado work-ing through an RNA-mediated signal transduction path-way. 8-NH2-Ado is known to target the cancer cellsthrough the inhibition of transcription as well as throughATP depletion (39). In our hands, common signal trans-duction outcomes of 8-NH2-Ado treatment were the inhi-bition of p21 transcription and the increase in LC3-II.

However, when we compared the induction of cell deathin MCF-7 cells and MDA-MB-231 cells, we observed theengagement of different signal transduction pathways. InMCF-7 cells, 8-NH2-Ado induced unequivocal apoptosisindependently of p53. However, inMDA-MB-231 cells, 8-NH2-Ado induced a signal transduction pathway towarddeath that was p53 independent and difficult to define.Inhibiting autophagy or necrosis did not block this deathpathway.

Important for future therapeutic consideration is thefact that 8-NH2-Ado inhibited breast cancer cell prolifer-ation and induced breast cancer cell death withoutrequiring the activation of wild-type p53. The ability of8-NH2-Ado to use RNA-dependent cell killing signaltransduction pathways is an excellent strategy for treatingcancers resistant to the current therapeutic options that

Figure 6. 8-NH2-Ado increases theautophagy markers in MDA-MB-231 cells but does not requireautophagy for cytotoxicity MCF-7and MDA-MB-231 cells weretreated with 15 mmol/L 8-NH2-Adofor 24 hours. A, Western blotanalysis was used to compare cellextracts fromMDA-MB-231, MCF-7, or MCF7.beclin 1 cells treatedwith 15 mmol/L 8-NH2-Ado for 24hours for PARP cleavage, Beclin 1,and LC3 and actin. B, MDA-MB-231 cells were treated with theautophagy inhibitor 3-MA alongwith the addition of 8-NH2-Ado andPARP was examined by Westernblot. C, the MTT assay was usedto determine MDA-MB-231mitochondrial activity following8-NH2-Ado with or without theaddition of 3-MA. D, phasecontrast microscopy at �20magnification showed no reversionof 8-NH2-Ado-induced death ofMDA-MB-231 cells with previoustreatment with 3-MA. E, Westernblot analysis was used to assessATG7 and PARP levels innontargeted siRNA or atg7 siRNAwith and without 8-NH2-Adotreatment in MDA-MB-231 cells.F, the MTT assay was used todetermine MDA-MB-231mitochondrial activity after 8-NH2-Ado with or without the addition ofnontargeted siRNA or atg7 siRNA.G, the structure of 3-MA wasobtained from chemicalbook.com.

Polotskaia et al.


depend on DNA damage. It is important to determinetherapeutics that can be equally effective on heteroge-neous tumors that have lost their functional p53 path-way, their DNA damage pathway, or their apoptoticpathway. The molecular mechanism of action of 8-NH2-Ado fits this paradigm and would increase the chancesof eradicating hard to kill cancers. We showed that the8-NH2-Ado inhibited the proliferation of 4 differentbreast cancer cell lines better than the commonly usedtherapeutics gemcitabine and etoposide. Moreover, thecytotoxicity of the drug was not influenced by thepresence of wild-type or mutant p53.MCF-7 cells are an example of cancer cellswith deficient

autophagy and apoptotic pathways but they were highlysensitive to 8-NH2-Ado. ThisMCF-7 cancer cell line iswelldocumented for beinghaploinsufficient for the autophagygene beclin 1 (19). Interestingly, increased expression ofBeclin 1 in MCF-7.beclin1 cells inhibited their apoptosisafter 8-NH2-Ado treatment but did not influence theability of the drug to block MCF-7.beclin1 proliferation.MDA-MB-231 cells undergo autophagic cell death (withan associated increase in Beclin 1 protein) when treatedwith an indole-3-carbinol metabolite (40); however, wesaw that 8-NH2-Ado–induced cell death of MDA-MB-231cells did not require autophagy. Inhibition of autophagyby 3-MA or knockdown of atg7 did not inhibit the MDA-MB-231 cell death. This indicates that 8-NH2-Adodoesnotinduce cell death by autophagy but rather that autophagyaccompanies the cell death.The efficacy of breast cancer cell killing by 8-NH2-Ado

regardless of the p53 status, or the estrogen receptorstatus, makes this nucleoside analogue an attractivechoice for patients who failed previous therapies. Mostrecently, 8-NH2-Ado was suggested as a therapeutic toinhibit BCR-ABL mRNA and protein levels in imatinib-resistant cancers (39). This combination treatment withimatinib and 8-NH2-Ado inhibits cell growth withoutincreasing Annexin positivity suggesting a mechanismother that apoptosis (39). On the basis of our breast cancerstudieswith 8-NH2-Ado,we conclude that this nucleosideanalog should be investigated further as a therapeuticoption for breast cancers that have failed other treatments.The RNA-directed nucleoside analogue 8-chloro-adeno-sine is currently in phase I clinical trials and as high as 500

nmol/L levels of drug are achieved in plasma (39). Thissuggests feasibility of using these drugs at the neededdosage in human patients. In addition, previous studiesshow that 8-NH2-Ado is not cytotoxic to normal lympho-cytes (13, 14). When we tested reduction mammoplastycells, they were sensitive to 8-NH2-Ado in the culturesetting but these cells were very difficult to grow andtherefore were not a good indicator of overall cytotoxicityto apatient (Polotskaia andBargonetti, unpublisheddata).The outcomes of 8-chloro-adenosine studies should helpto pave the way for the consideration of 8-NH2-Ado as anagent to treat aggressive breast cancer.

Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.

Authors' ContributionsConception and design: A. Polotskaia, S. Hoffman, N. L. Krett, M.Shanmugam, J. BargonettiDevelopment of methodology: A. Polotskaia, M. Shanmugam, J.BargonettiAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): A. Polotskaia, J. BargonettiAnalysis and interpretation of data (e.g., statistical analysis, biostatis-tics, computational analysis): A. Polotskaia, S. Hoffman, N. L. Krett, S. T.Rosen,J. BargonettiWriting, review, and/or revision of the manuscript: A. Polotskaia, S.Hoffman, S. T. Rosen, J. BargonettiAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): A. Polotskaia, J. BargonettiStudy supervision: J. Bargonetti

AcknowledgmentsThe authors thank Beth Levine for theMCF7.beclin 1 clones andGuXiao

for discussions on the manuscript.

Grant SupportThis work was supported by a grant from the The Breast Cancer

Research Foundation and NIH SCORE Award 1SC1CA137843 to J. Bar-gonetti. It was also made possible in part by an infrastructure award toHunterCollege,Grant no.RR003037 from theNationalCenter forResearchResources (NCRR), a component of the NIH. Its contents are solely theresponsibility of the authors and do not necessarily represent the officialviews of NCRR or NIH.

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

Received January 29, 2012; revisedAugust 20, 2012; acceptedAugust 30,2012; published OnlineFirst September 12, 2012.

References1. Ewald B, Sampath D, Plunkett W. Nucleoside analogs: molecular

mechanisms signaling cell death. Oncogene 2008;27:6522–37.2. Feng L, AchantaG, PelicanoH, ZhangW, PlunkettW, HuangP. Role of

p53 in cellular response to anticancer nucleoside analog-inducedDNAdamage. Int J Mol Med 2000;5:597–604.

3. VousdenKH, LaneDP. p53 in health and disease. Nat RevMol Cell Biol2007;8:275–83.

4. Nigro JM, Baker SJ, Presinger C, Jessup JM, Hostetter R, Cleary K,et al. Mutations in the p53 gene occur in diverse human tumor types.Nature 1989;342:705–8.

5. Lacroix M, Toillon RA, Leclercq G. p53 and breast cancer, an update.Endocr Relat Cancer 2006;13:293–325.

6. Rivera E, Gomez H. Chemotherapy resistance in metastatic breastcancer: the evolving role of ixabepilone. Breast Cancer Res 2010;12Suppl 2:S2.

7. Christophorou MA, Ringshausen I, Finch AJ, Swigart LB, Evan GI. Thepathological response to DNA damage does not contribute to p53-mediated tumour suppression. Nature 2006;443:214–7.

8. Hastak K, Alli E, Ford JM. Synergistic chemosensitivity of triple-negative breast cancer cell lines to PARP inhibition, gemcitabine andcisplatin. Cancer Res 2010;70:7970–80.

9. Alli E, Sharma VB, Hartman AR, Lin PS, McPherson L, Ford JM.Enhanced sensitivity to cisplatin and gemcitabine in Brca1-deficientmurine mammary epithelial cells. BMC Pharmacol 2011;11:7.



10. Anders CK, Winer EP, Ford JM, Dent R, Silver DP, Sledge GW, et al.Poly (ADP-Ribose) polymerase inhibition: "targeted" therapy for triple-negative breast cancer. Clin Cancer Res 2010;16:4702–10.

11. Stellrecht CM, Ayres M, Arya R, Gandhi V. A unique RNA-directednucleoside analog is cytotoxic to breast cancer cells and depletescyclin E levels. Breast Cancer Res Treat 2010;121:355–64.

12. Frey JA, Gandhi V. 8-Amino-adenosine inhibits multiple mechanismsof transcription. Mol Cancer Ther 2010;9:236–45.

13. Krett NL, Davies KM, Ayres M, Ma C, Nabhan C, Gandhi V, et al. 8-Amino-adenosine is a potential therapeutic agent for multiple myelo-ma. Mol Cancer Ther 2004;3:1411–20.

14. DennisonJB, ShanmugamM,AyresML,Qian J, Krett NL,Medeiros LJ,et al. 8-Aminoadenosine inhibits Akt/mTORandErk signaling inmantlecell lymphoma. Blood 2010;116:5622–30.

15. Chen LS, Du-Cuny L, Vethantham V, Hawke DH, Manley JL, ZhangS, et al. Chain termination and inhibition of mammalian poly(A)polymerase by modified ATP analogues. Biochem Pharmacol2010;79:669–77.

16. ShanmugamM,McBrayer SK, Qian J, Raikoff K, AvramMJ, Singhal S,et al. Targeting glucose consumption and autophagy in myeloma withthe novel nucleoside analogue 8-aminoadenosine. J Biol Chem2009;284:26816–30.

17. Ghias K, Ma C, Gandhi V, Platanias LC, Krett NL, Rosen ST. 8-Amino-adenosine induces loss of phosphorylation of p38 mitogen-activatedprotein kinase, extracellular signal-regulated kinase 1/2, and Aktkinase: role in induction of apoptosis in multiple myeloma. Mol CancerTher 2005;4:569–77.

18. Brekman A, Singh KE, Polotskaia A, Kundu N, Bargonetti J. A p53-independent role of Mdm2 in estrogen-mediated activation of breastcancer cell proliferation. Breast Cancer Res 2011;13:R3.

19. Liang XH, Jackson S, SeamanM, Brown K, Kempkes B, Hibshoosh H,et al. Induction of autophagy and inhibition of tumorigenesis by beclin1. Nature 1999;402:672–6.

20. Boamah EK, White DE, Talbott KE, Arva NC, Berman D, Tomasz M,et al. Mitomycin-DNA adducts induce p53-dependent and p53-inde-pendent cell death pathways. ACS Chem Biol 2007;2:399–407.

21. Adorno M, Cordenonsi M, Montagner M, Dupont S, Wong C, Hann B,et al. A mutant-p53/Smad complex opposes p63 to empowerTGFbeta-induced metastasis. Cell 2009;137:87–98.

22. Neve RM, Chin K, Fridlyand J, Yeh J, Baehner FL, Fevr T, et al. Acollection of breast cancer cell lines for the studyof functionally distinctcancer subtypes. Cancer Cell 2006;10:515–27.

23. Marine JC, Lozano G. Mdm2-mediated ubiquitylation: p53 andbeyond. Cell Death Differ 2010;17:93–102.

24. Shieh SY, Ikeda M, Taya Y, Prives C. DNA damage-induced phos-phorylation of p53 alleviates inhibition byMDM2.Cell 1997;91:325–34.

25. Brosh R, Rotter V. When mutants gain new powers: news from themutant p53 field. Nat Rev Cancer 2009;9:701–13.

26. Tafani M, Schito L, Anwar T, Indelicato M, Sale P, Di Vito M, et al.Induction of autophagic cell death by a novel molecule is increased byhypoxia. Autophagy 2008;4:1042–53.

27. Augustin S, Berard M, Kellaf S, Peyri N, Fauvel-Lafeve F, Legrand C,et al. Matrix metalloproteinases are involved in both type I (apoptosis)and type II (autophagy) cell death induced by sodium phenylacetate inMDA-MB-231 breast tumour cells. Anticancer Res 2009;29:1335–43.

28. Kroemer G, Levine B. Autophagic cell death: the story of a misnomer.Nat Rev Mol Cell Biol 2008;9:1004–10.

29. MelendezA,Neufeld TP. The cell biology of autophagy inmetazoans: adeveloping story. Development 2008;135:2347–60.

30. Tanida I, Ueno T, Kominami E. LC3 conjugation system in mammalianautophagy. Int J Biochem Cell Biol 2004;36:2503–18.

31. Mizushima N, Klionsky DJ. Protein turnover via autophagy: implica-tions for metabolism. Annu Rev Nutr 2007;27:19–40.

32. 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.

33. Xue LY, Chiu SM, Oleinick NL. Atg7 deficiency increases resistance ofMCF-7 human breast cancer cells to photodynamic therapy. Autop-hagy 2010;6:248–55.

34. LevineB.Cell biology: autophagy andcancer. Nature 2007;446:745–7.35. Yue Z, Jin S, Yang C, Levine AJ, Heintz N. Beclin 1, an autophagy gene

essential for early embryonic development, is a haploinsufficient tumorsuppressor. Proc Natl Acad Sci U S A 2003;100:15077–82.

36. Kondo Y, Kondo S. Autophagy and cancer therapy. Autophagy2006;2:85–90.

37. Chen N, Debnath J. Autophagy and tumorigenesis. FEBS Lett 2010;584:1427–35.

38. LockR, RoyS, KenificCM, Su JS, Salas E, Ronen SM, et al. Autophagyfacilitates glycolysis during Ras-mediated oncogenic transformation.Mol Biol Cell 2011;22:165–78.

39. Pillai RN, Chen LS, AyresML, Nowak BJ, ThomasMW, Shpall EJ, et al.Multifaceted actions of 8-amino-adenosine kill BCR-ABL positivecells. Leuk Lymphoma 2012;53:2024–32.

40. Vanderlaag K, Su Y, Frankel AE, Burghardt RC, Barhoumi R, Chada-lapaka G, et al. 1,1-Bis(30-indolyl)-1-(p-substituted phenyl)methanesinduce autophagic cell death in estrogen receptor negative breastcancer. BMC Cancer 2010;10:669.

Polotskaia et al.


8-amino-adenosineactivatesp53-independentcelldeathof metastatic breast … · breast cancer cell...

Documents