Suppression of survivin expression inhibits in vivo tumorigenicity and angiogenesis in gastric cancer

[CANCER RESEARCH 63, 7724–7732, November 15, 2003]

Suppression of Survivin Expression Inhibits in Vivo Tumorigenicity andAngiogenesis in Gastric Cancer

Shui Ping Tu,1,2 Xiao Hua Jiang,1,2 Marie C. M. Lin,3 Jian Tao Cui,2 Yi Yang,2 Ching Tung Lum,3 Bing Zou,1,2

Yan Bo Zhu,1 Shi Hu Jiang,1 Wai Man Wong,2 Annie On-On Chan,2 Man Fung Yuen,2 Shiu Kum Lam,2

Hsiang Fu Kung,3 and Benjamin Chun-Yu Wong2

1Department of Gastroenterology, Rui-jin Hospital, Shanghai Second Medical University, Shanghai, People’s Republic of China, and 2Department of Medicine and 3Institute ofMolecular Biology, University of Hong Kong, Hong Kong Special Administration Region, People’s Republic of China

ABSTRACT

Survivin plays an important role in cancer development. We aim toshow here that suppression of survivin expression or function by antisenseand dominant-negative (DN) mutant can inhibit gastric cancer carcino-genesis and angiogenesis in vivo. Plasmid constructs expressing survivinantisense and DN mutant replacing the cysteine residue at amino acid 84with alanine (Cys84Ala) were prepared and introduced into BCG-823 andMKN-45 gastric cancer cells to establish stable transfectants. We showedthat both antisense and DN mutant stable transfectants exhibited abnor-mal morphology, with decreased cell growth and increased rate of spon-taneous apoptosis and mitotic catastrophe. Furthermore, in nude micexenografts, these cells exhibited decreased de novo gastric tumor forma-tion and reduced development of angiogenesis. Results from these studiesstrongly suggest that survivin is a promising target for gastric cancertreatment.

INTRODUCTION

Survivin, a structurally unique IAP4 family protein, is expressed inmitosis in a cell cycle-dependent fashion and localized to componentsof the mitotic apparatus (1). It is potentially involved in both theinhibition of apoptosis and control of cell division (2, 3). Survivin isfound in most human cancers but is either undetectable or expressedat a very low level in differentiated adult tissues (4). In most cancers,expression of survivin correlated with reduced apoptotic index, poorprognosis, and increased risk of recurrence (5–8). Antisense targetingof survivin results in the dysregulation of mitotic spindle checkpointas well as defects in microtubule assembly and function that lead tomitotic catastrophe (9–12). Mitotic catastrophe is the form of celldeath that results from aberrant mitosis, which is characterized bysupernumerary centrosomes, failure of cytokinesis, and a significantincrease in the percentage of abnormal nuclei, the appearance ofwhich includes multiple multilobulated nuclei and abnormally large-sized nuclei (13). A survivin DN Thr34Ala mutant, with the threonineresidue at amino acid 34 changed to alanine and the phosphorylationsite for p34cdc2-cyclin B1 abolished, has also been shown to causeapoptosis and inhibit tumor formation and tumor growth (14–16).These unique features of survivin make it a promising target forcancer therapy. However, most of the studies regarding the role ofsurvivin in apoptosis and cell division are performed using the tran-

sient expression method. Furthermore, the in vivo mechanism oftargeting survivin in cancer treatment is not fully understood (17, 18).

It has been shown that survivin is up-regulated in angiogenicallystimulated endothelium in vitro and in vivo but undetectable in qui-escent endothelial cells (19, 20). Angiogenic agents, such as vascularendothelial growth factor and basic fibroblast growth factor, inducesurvivin expression in endothelial cells (21). Because tumor angio-genesis depends on endothelial viability, it is conceivable that target-ing survivin may favor apoptotic involution of newly formed bloodvessels and indirectly inhibit tumor formation. However, whetherknock-down of survivin function in tumor cells will affect angiogen-esis in vivo remains unknown.

Gastric cancer is one of the most common malignant tumors world-wide. Studies show that 35–82.6% of gastric cancers expressed sur-vivin, and survivin expression correlated with poor survival of pa-tients (22–24). Moreover, up-regulation of survivin is found in gastriccancer cell lines after treatment with cytotoxic drugs, indicating thatsurvivin contributes to chemoresistance in gastric cancer (25). In thisstudy, we aimed to study the effect of constitutive suppression ofsurvivin on gastric cancer carcinogenesis using a DN mutant ofsurvivin, replacing the cysteine residue at amino acid 84 with alanine(Cys84Ala), that binds to the mitotic apparatus and displaces wild-type survivin from polymerized microtubules (19). We establishedstable cell lines expressing Sur-AS cDNA or DN Sur-Mut (Cys84Ala)and investigated the effect of targeting survivin in gastric cancertreatment. Our results showed that suppression of survivin inducedboth apoptosis and mitotic catastrophe in vitro and in vivo in gastriccancer epithelial cells. The antisense or mutant (Cys84Ala)-mediatedsuppression of survivin function also inhibited tumor formation andangiogenesis in gastric cancer xenograft model in vivo.

MATERIALS AND METHODS

Cell Lines and Cell Culture. We selected two human gastric adenocarci-noma cell lines for this study. BCG-823 (Beijing Institute of Cancer Research,Beijing, China) is an adherent, poorly differentiated, human gastric adenocar-cinoma cell line with mutant p53 (26–28), and the MKN-45 cell line (RIKEN,the Institute of Physical and Chemical Research, Cell Bank, Ibaraki, Japan)expresses wild-type p53; however, it has p21, p16 homozygous deletion, andp27 rearrangement (29). Both cell lines are gastric cancer epithelial cells andgrow as adherent cells in RPMI 1640 (Life Technologies, Inc., Grand Island,NY) containing 10% FCS (Life Technologies, Inc.), 2 mM L-glutamine (Bio-Whittaker, Walkersville, MD), 100 units/ml penicillin, and 100 �g/ml strep-tomycin (BioWhittaker). All cell lines were maintained at 37°C in a humidifiedincubator with an atmosphere of 5% CO2. Cisplatin and 5-fluorouracil (Phar-macia & Upjohn, Ltd. Corp., Chatswood, Australia) were solublized in DMSOto concentrations of 2 and 10 �g/ml, respectively, and stored at 4°C.

Construction of Sur-AS and DN Mutant Plasmids. Total RNA wasextracted from MKN-45 cells using Trizol reagent (Life Technologies, Inc.).cDNA coding survivin was generated by reverse transcription-PCR using thesurvivin forward primer (5�-GGGAATTCATGGGTGCCCCGACGTTGCC-3�) and reverse primer (5�-CTCTCGAGTCAATCCATGGCAGCCAGCT-3�;Genset Singapore Biotech, Pte Ltd., Singapore). The PCR product was insertedinto the EcoRI and XhoI sites of pcDNA3(�) and pcDNA3(�) (Invitrogen,

Received 4/8/03; revised 7/31/03; accepted 9/12/03.Grant support: Research Grant Council earmarked Grant HKU 7309/01M of the

Hong Kong Special Administration Region, the Simon KY Lee Gastroenterology Re-search Fund, Queen Mary Hospital and Gastroenterogical Research Fund, the Universityof Hong Kong, Hong Kong Special Administration Region, People’s Republic of China.

The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

Requests for reprints: Benjamin Chun-Yu Wong, Department of Medicine, Univer-sity of Hong Kong, Queen Mary Hospital, Hong Kong, People’s Republic of China.Phone: 852-2855-4541; Fax: 852-2872-5828; E-mail: [email protected].

4 The abbreviations used are: IAP, inhibitor of apoptosis protein; DN, dominant-negative; Sur-AS, survivin antisense; Sur-Mut, survivin mutant; MTT, 3-(4,5-dimethyl-thiazol-2-yl)-2,5-diphenyltetrazolium bromide; TUNEL, terminal deoxynucleotidyltrans-ferase-mediated nick end labeling; MVD, microvessel density; PI, propidium iodide.

7724

Research. on November 27, 2015. © 2003 American Association for Cancercancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

Groningen, the Netherlands) to generate pcDNA3-surivin and pcDNA3(�)-Sur-AS plasmids, respectively.

We used an overlap extension PCR to construct a DN mutant of survivinusing pcDNA3-survivin plasmid as the template. The 5�-flanking forwardprimer and 3�-flanking reverse primer were as described above. For Sur-MutCys84Ala, the forward primer was 5�-ATTCGTCCGGCGCCGCTTTCCTTTCTG-3�, and the reverse primer was 5�-CAGAAAGGAAAGCGGCGCCG-GACGAAT-3�, which produced a TGC-to-GCC substitution at nucleotide251-253 and resulted in replacement of the cysteine residue at amino acid 84with an alanine. The fragment containing Sur-Mut was cloned in the EcoRI andXhoI site of the pcDNA3 vector, generating pcDNA3-Sur-Mut (Cys84Ala)plasmid. The precision of all of the constructs was confirmed by sequencing.

Establishment of BCG-823 and MKN-45 Stable Transfectants Express-ing Su-AS and Sur-Mut. For transfection experiments, BCG-823 andMKN-45 cells were plated into 6-well plates (3 � 105 cells/well) 18 h beforetransfection. The cells were transfected with 4 �g/well of empty pcDNA3vector, pcDNA3-Sur-AS, or pcDNA3-Sur-Mut (Cys84Ala) plasmid using Li-pofectAMINE 2000 (Life Technologies, Inc., Grand Island, NY) according tothe manufacturer’s instructions. Forty-eight h after transfection, the cells werepassaged at 1:15 (v/v) and cultured in medium supplemented with Geneticin

(G418) at 1000 �g/ml for 4 weeks. Stably transfected clones were picked andmaintained in medium containing 400 �g/ml G418 for additional studies.

Assay of Anchorage-Dependent Cell Growth. Parent cells and cells sta-bly expressing pcDNA3, pcDNA3-Sur-AS, or pcDNA3-Sur-Mut (1 � 105)were seeded into 6-well plates. Cells from triplicate wells were collected everyother day. Cell numbers were determined using a Coulter counter (CoulterElectronics, Miami, FL). The number of cells per well is reported as theaverage � SD at the indicated number of days after plating.

Assay for Anchorage-Independent Cell Growth. Cells (1 � 104) wereplated in complete culture medium containing 0.3% agar on top of 0.6% agarin the same medium. All dishes were incubated at 37°C in a humidifiedatmosphere of 5% CO2. Colonies were scored 16 days after plating, fixed with70% ethanol, stained with Coomassie Blue, and counted under a dissectionmicroscope. Colonies containing �50 cells were considered viable.

Acridine Orange and PI Staining. Cells were fixed with 4% formalin/PBS, stained with 10 �g/ml acridine orange (Sigma, St. Louis, MO) or 200�g/ml PI, and visualized under a fluorescence microscope. Apoptotic cellswere defined as cells showing cytoplasmic and nuclear shrinkage, chromatincondensation, or fragmentation morphologically. At least 300 cells/field werecounted to determine the apoptotic index or the number of aberrant nuclei.

Fig. 1. A, the expression of survivin protein in the parental vector transfectants, Sur-AS transfectants, and Sur-Mut (Cys84Ala) transfectants of BCG-823 and MKN-45 cells. B, theexpression of other IAP family proteins in BCG-823 cells evaluated by immunoblotting assay. The blots were reacted with specific antibodies recognizing survivin, XIAP, c-IAP-1,c-IAP-2, livin, and �-actin, respectively. C, the BCG-823 transfectants were grown to confluence and either photographed with a phase-contrast microscope (top panels) or fixed with4% paraformalin, stained with 4�,6-diamidino-2-phenylindole, and photographed under fluorescence microscopy (bottom panels). Arrows indicate abnormal morphological cells(original magnification, �200).

7725

SUPPRESSION OF SURVIVIN EXPRESSION IN GASTRIC CANCER



MTT Assay. Cytotoxicity was measured by MTT assay. Cells growingexponentially were plated onto 96-well plates at a density of 1000 cells/wellfor 24 h. The cells were then treated with different concentrations of drugs for96 h. One hundred �l of MTT stock solution (1 mg/ml) were added to eachwell, and the cells were further incubated at 37°C for 4 h. The supernatant wasreplaced with isopropyl alcohol to dissolve formazan production. The absorb-ance at wavelength 595 nm was measured with a micro-ELISA reader (Bio-Rad, Hercules, CA). The ratio of the absorbance of treated cells relative to thatof the control cells was calculated and expressed as a percentage of cell death.

Flow Cytometry. Cells were collected and fixed with ice-cold 70% ethanolin PBS and stored at –4°C until use. After resuspension, cells were incubatedwith 100 �l of RNase I (1 mg/ml) and 100 �l of PI (400 �g/ml) at 37°C andanalyzed by flow cytometry (Coulter, Luton, United Kingdom). The cell cyclephase distribution was calculated from the resultant DNA histogram usingMulticycle AV software (Phoenix Flow Systems, San Diego, CA). Cells witha subdiploid DNA content were considered apoptotic cells.

Immunofluorescence. Cells were grown on a 2-well glass chamber slide(Nunc, Naperviller, IL) and transfected with plasmids, as described previously.Cells were fixed with 2% formaldehyde for 10 min and permeabilized with0.5% NP40 in PBS (�-tubulin) or fixed and permeabilized in ice-cold methanolfor 20 min (�-tubulin). Antibodies to �-tubulin (clone DM1A) and �-tubulin(clone GTU-88), and FITC-conjugated goat antimouse antibodies were pur-chased from Sigma. Antibodies were used at a 1:100 dilution for tubulins.

Nuclei were stained with 1 �g/ml Hoechst 22358, and cells were analyzedusing a Zeiss Axioscop fluorescence microscope.

In Situ Detection of Apoptotic Cells by TUNEL Assay. Tumor tissuesfrom xenografts were excised and formalin-fixed immediately after resection.TUNEL staining was carried out using the ApoAlert DNA fragmentation assaykit (Clontech Corp., Palo Alto, CA) according to the manufacturer’s instruc-tions. Apoptotic cells exhibit strong nuclear green fluorescence. The percent-age of apoptotic cells was assessed in 10 randomly selected fields viewed at�40 magnification. The apoptotic index was calculated as the number ofapoptotic cells/total number of nucleated cells �100%.

Western Blot Analysis. Cells were lysed with lysis buffer [50 mM Tris-HCl (pH 7.5), 250 mM NaCl, 0.1% NP40, and 5 mM EGTA containing 50 mM

sodium fluoride, 60 mM �-glycerol-phosphate, 0.5 mM sodium vanadate,0.1 mM phenylmethylsulfonyl fluoride, 10 �g/ml aprotinin, and 10 �g/mlleupeptin]. Protein samples were electrophoresed in a 10% denaturing SDS geland transferred to Immobilon-P membrane (Millipore, Bedford, MA). Theblots were incubated with specific primary antibodies, reacted with a peroxi-dase-conjugated secondary antibody (Santa Cruz Biotechnology, Santa Cruz,CA), and finally visualized by enhanced chemiluminescence (Amersham,Piscataway, NJ). Polyclonal antibodies recognizing survivin (2 �g/ml) andlivin (2 �g/ml) were purchased from Alpha Diagnostic International, Inc. (SanAntonio, TX); XIAP (2 �g/ml) monoclonal antibody was obtained fromStressGen Biotechnologies Corp. (Victoria, British Columbia, Canada); and

Fig. 2. Anchorage-dependent and anchorage-independentgrowth in single and pooled BCG-823 and MKN-45 stable trans-fectants. A, anchorage-dependent growth was counted in BCG-823parent cells (F), BCG-823/vector (E), single clone of BCG-823/Sur-AS1 (Œ), BCG-823/Sur-AS2 (‚), BCG-823/Sur-AS3 (f), andpooled BCG-823/Sur-Mut transfectants (�). B, anchorage-depen-dent growth was counted in BCG-823 parent cells (F), BCG-823/vector (E), single BCG-823/Sur-Mut1 (Œ), BCG-823/Sur-Mut2(‚), BCG-823/Sur-Mut3 (f), and pooled BCG-823/Sur-Mut trans-fectants (�). Cells from triplicate wells were collected every otherday. Data represent mean � SE of three independent experiments.�, P � 0.01; #, P � 0.001 (compared with control vector trans-fectants). Anchorage-independent growth was analyzed by softagar assay in BCG-823 parent cells and single and pooled BGC-823/Sur-AS and BCG-823/Sur-Mut transfectants (C and D) and inMKN-45 parent cells and single and pooled MKN-45/Sur-AS andMKN-45/Sur-Mut transfectants (E and F). Data represent themean � SE of three independent experiments. �, P � 0.001compared with BCG-823/vector.

7726




c-IAP1 and c-IAP2 polyclonal antibodies, poly(ADP-ribose) polymerase,�-actin, bcl-2, bax, bcl-xl, bcl-xs, and cytochrome c monoclonal antibodieswere purchased from Santa Cruz Biotechnology.

Immunohistochemistry. The expression of survivin and Ki67 in tumortissues was detected with the DAKO labeled streptavidin-biotin kit (DAKO,Carpinteria, CA) according to the manufacturer’s instructions. Briefly, slideswere boiled in 10 mM citrate buffer (pH 6; Bio Genex, San Ranmon, CA) forantigen retrieval and incubated with antisurvivin polyclonal antibody (1:100;Alpha Diagnostic International, Inc.), Ki67 polyclonal antibody (1:100; Neo-Markers, Fremont, CA), and CD31 (PECAM-1) polyclonal antibody (1:50;PharMingen, San Diego, CA), respectively, followed by biotinylated anti-IgGantibody (1:200; DAKO) and streptavidin-biotinylated-complex/horseradishperoxidase (DAKO). The sections were counterstained with hematoxylin andindependently evaluated by two blinded investigators. Survivin and Ki67staining were recorded as the ratio of positively stained cells to all tumor cellsin five different areas at �200 magnification. MVD was evaluated accordingto method described previously (30). MVD was the average of the vesselcounts obtained in the three sections. Areas of the highest neovascularizationwere chosen, and microvessel counting was performed at �200 magnificationin three chosen fields. Any immunoreactive endothelial cell or endothelial cellcluster that had been distinctly separated from adjacent microvessels wasconsidered a single countable vessel. The results regarding angiogenesis ineach tumor were expressed as the absolute number of vessels/0.74 mm2 (�200field). In all assays, matched isotype control antibodies were used and found tobe unreactive in all cases.

Gastric Cancer Xenograft Animal Model and Tumorigenicity Assay.Female BALB/c nude mice, 5–6 weeks old, were bred in the Animal Labo-ratory Unit, the University of Hong Kong, Hong Kong. Institutional guidelineswere followed in handling the animals. The tumors were established by s.c.injection of 1 � 106 BCG-823 cells, BCG-823/vector, BCG-823/Sur-AS, and

BCG-823/Sur-Mut (Cys84Ala) transfectants/0.2 ml PBS into the right flanksof the mice. Tumor sizes were determined by measuring two diametersperpendicular to each other with a caliper every 3 days. Tumor volume (V) wasestimated by using the equation V � 4/3� � L/2 � (W/2)2, where L is themid-axis length, and W is the mid-axis width. Four mice were included in eachgroup in all experiments, and each experiment was performed twice. Animalswere sacrificed 30 days after implantation. Tumor tissues were removed, fixedin formalin or zinc fixative, embedded in paraffin, and subjected to H&E,TUNEL, or immunohistochemical staining. The protocol was approved by theCommittee on the Use of Live Animals in Teaching and Research, theUniversity of Hong Kong, Hong Kong.

Statistical Analysis. Data were expressed as the means of at least threedifferent experiments � SD. The results were analyzed by Student’s t test, andP �0.05 was considered statistically significant.

RESULTS

Characterization of Stable Transfectants Expressing Sur-ASand Sur-Mut. We established BCG-823 and MKN-45 stable trans-fectants with either control vector pcDNA3, Sur-AS plasmid, orSur-Mut plasmid. Eight clones from each transfection were selectedand analyzed by Western blot to determine the survivin proteinexpression, and three clones were selected for expansion and addi-tional studies. As shown in Fig. 1A, the level of survivin protein inpooled BCG-823 Sur-AS transfectants was reduced by �90%, al-though pooled Sur-Mut (Cys84Ala) transfectants expressed 2.53-foldsurvivin protein compared with that expressed by BCG-823 parentalcells (Fig. 1A). In contrast, no change in the protein level of other IAP

Fig. 3. Sensitization to cytotoxic drug and serum depletioninduced apoptosis. A and B, cytotoxicity assay. BCG-823 parentcells (E), BCG-823/vector (Œ), BCG-823/Sur-AS (‚), andBCG-823/Sur-Mut transfectants (F) were plated into 96-wellplates for 24 h. Cells were then treated with indicated dose ofdrugs for 96 h. Cytotoxicity was measured by MTT assay.Apoptosis assay: BCG-823 (C) and MKN-45 (D) parental cellsand stable transfectants were incubated with 2 �g/ml cisplatinor 10 �g/ml 5-fluorouracil for 48 h. BCG-823 (E) and MKN-45(F) parental cells and stable transfectants were incubated inmedium with 10% fetal calf serum for 24 h and switched toserum-free medium for another 48 h. The percentage of apo-ptotic cells was measured by fluorescence-activated cell sorter.These results represent the mean � SE of three independentexperiments. �, P � 0.01 compared with parental cells andcontrol transfectants. �, P � 0.001 compared with untreatedcells. #, P � 0.01 compared with those cultured in 10% serum.Parental cells, �; control vector transfectants, u; Sur-AS trans-fectants, f; Sur-Mut transfectants, o.

7727




family genes was detected in all of the stable clones isolated fromeither BCG-823 (Fig. 1B) or MKN-45 (data not shown) cells. Fur-thermore, the BCG-823/Sur-AS and BCG-823/Sur-Mut (Cys84Ala)transfectants exhibit aberrant morphology with larger and flatter cellsobserved under phase-contrast microscope (Fig. 1C, top panels). Asdemonstrated by 4�,6-diamidino-2-phenylindole staining, BCG-823/Sur-AS and BCG-823/Sur-Mut (Cys84Ala) transfectants exhibitedflattened and giant nuclei, whereas parental cells and BCG-823/vector-transfected cells showed normal morphology (Fig. 1C, bottompanels). Similar results were also obtained from the MKN-45 stabletransfectant (data not shown).

Inhibition of Cell Growth in Stable Transfectants. As shown inFig. 2, single transfectant BCG-823/Sur-AS1, BCG-823/Sur-AS2, andBCG-823/Sur-AS3 and pooled BCG-823/Sur-AS-P transfectants orsingle transfectant BCG-823/Sur-Mut1, BCG-823/Sur-Mut2, andBCG-823/Sur-Mut3 and pooled BCG-823/Sur-Mut (Cys84Ala) cellshad significant decreases in cell number compared with BCG-823/

vector-transfected and parental cells 5 days after plating (P � 0.01).By day 7, the number of cells was reduced by 64 � 7% and 78 � 9%in BCG-823/Sur-AS-P and BCG-823/Sur-Mut (Cys84Ala)-P transfec-tants compared with parental cells (P � 0.001; Fig. 2, A and B).Similar results were obtained in MKN-45/Sur-AS and MKN-45/Sur-Mut transfectants (data not shown). Furthermore, single and pooledBCG-823/Sur-AS and BCG-823/Sur-Mut (Cys84Ala) transfectantsformed significantly fewer and smaller colonies in soft agar (Fig. 2, Cand D). Colony forming efficiencies in BCG-823/Sur-AS-P and BCG-823/Sur-Mut (Cys84Ala)-P cells were reduced by 80 � 5% and90 � 5%, respectively, as compared with parental cells (P � 0.001;Fig. 2B). Similar results were obtained in single and pooled MKN-45/Sur-AS and MKN-45/Sur-Mut transfectants (Fig. 2, E and F).

Stable Transfectants Were More Susceptible to ProapoptoticStimuli. To avoid unintended differences related to gene insertionrather than to the expression of Sur-AS or Sur-Mut, we chose pooledclones to perform further study. We treated BCG-823/Sur-AS and

Fig. 4. Mitotic catastrophe in gastric cancercells. A, nuclear morphology and centrosomes inparental cells and stable transfectants. Cells werestained with an antibody to �-tubulin for spindlecomponents or �-tubulin for centrosomes. Cellswere analyzed by fluorescence microscopy (B, C,and top panels of A) or confocal laser-scanningmicroscope (D, E, and middle and bottom panels ofA). Photographs were from a representative exper-iment. B, representative normal nuclei. C, repre-sentative abnormal larger and multilobed nucleiand aberrant mitoses. D, representative chromo-some fusions. E, representative DNA bridges.Original magnification: �200 in top panels of Aand �320 in middle panels of A. Scale bar, bottompanels of A and B–E, 10 �m. F, quantification ofcells in aberrant nuclei in stable transfectants. G,quantification of cells in abnormal mitoses in sta-ble transfectants. Data represent the mean of threeindependent experiments. An average of 150–300cells were counted for each determination. �,P � 0.001 compared with control vector transfec-tants.

7728




BCG-823/Sur-Mut transfectants with serum deprivation and cytotoxicdrugs. As shown in Fig. 3, MTT assay showed that BCG-823 cellstransfected with Sur-AS or Sur-Mut plasmid were more sensitive tocisplatin and 5-fluorouracil than parental and vector-transfected cells(Fig. 3, A and B). The number of apoptotic cells induced by cisplatinand 5-fluorouracil increased to about 2.5–3-fold in BCG-823/Sur-ASand BCG-823/Sur-Mut compared with their control cells (P � 0.01;Fig. 3C). Similar results were obtained in MKN-45/Sur-AS andMKN-45/Sur-Mut stable transfectants (Fig. 3D). Furthermore, bothBCG-823/Sur-AS and BCG-823/Sur-Mut transfectants underwentspontaneous apoptosis more readily than parental and control vectorcells in the presence or absence of serum (P � 0.01; Fig. 3E). Similarresults were obtained in MKN-45/Sur-AS and MKN-45/Sur-Mut sta-ble transfectants (Fig. 3F).

Stable Transfectants Encountered Mitotic Catastrophe. We se-lected the BCG-823/Sur-AS and BCG-823/Sur-Mut transfectants toevaluate the morphological changes. Nuclei and centrosomes werestained with an antibody recognizing �-tubulin for spindle compo-

nents or �-tubulin for centrosomes. Cells were then analyzed byfluorescence microscopy. As shown in Fig. 4A, BCG-823/Sur-AS andBCG-823/Sur-Mut cells showed abnormal nuclei; in terms of mor-phology, they were micronucleated, multilobulated, and of abnor-mally large size as compared with parent and BCG-823/vector trans-fectants. Furthermore, BCG-823/Sur-AS and BCG-823/Sur-Mut cellshad significant increases in the percentage of aberrant chromosomes,such as chromatin bridges (Fig. 4, A and E, arrowhead) and chromo-some fusion (Fig. 4, A and D, small arrow), which characterized theexit from mitosis without completing chromosome segregation orcytokinesis. Percentages of aberrant nuclei in BCG-823/Sur-AS andBCG-823/Sur-Mut cells were 28 � 9% and 30 � 8%, respectively(Fig. 4F). Percentages of abnormal mitoses in BCG-823/Sur-AS andBCG-823/Sur-Mut cells were 17 � 3% and 19 � 4%, respectively(Fig. 4G).

Constitutive Suppression of Survivin Inhibited Tumor Forma-tion in Vivo in Nude Mice Xenografts. We then asked whetherstable transfectants in which the function of survivin has been

Fig. 5. Suppression of gastric tumor growth invivo. A, tumor growth curve. Parental BCG-823cells and stable transfectants (1 � 106) were in-jected s.c. into BALB/c nude mice, and tumorswere monitored every 3 days. Each point repre-sents the mean tumor size (as measured by threeperpendicular diameters). Each bar represents themean tumor size with 95% confidence intervals foreight mice. �, P � 0.001 compared with those fromcontrol transfectants. B, tumors in nude mice.BCG-823/vector transfectants (left panel) andBCG-823/Sur-Mut (Cys84Ala) transfectants (rightpanel) were injected s.c. into BALB/c nude mice.Photographs were taken from representative mice 4weeks after injection. C, tumors were excised, andprotein expression of survivin was determined byWestern blotting (left panel) and immunohisto-chemical staining (right panel) in tumor derivedfrom BCG-823, BCG-823/vector, and BCG-823/Sur-AS transfectants.

7729




knocked down could have reduced tumorigenicity in nude mice xe-nografts. We selected the BCG-823/Sur-AS and BCG-823/Sur-Muttransfectants to perform the in vivo study because BCG-823 cells havestronger tumorigenicity in nude mice than the other gastric cancer celllines used.5 In all mice that received injection of control vector-transfected cells, the cells readily formed localized palpable tumorscomprised of malignant cells that stained positively for survivin(Fig. 5, B and C). Tumors appeared within 1 week in all of the animalsof the control group (Fig. 5A). Two of eight (75%) mice that receivedinjection of the BCG-823/Sur-AS cells developed palpable tumorswith considerably smaller size starting at day 22 (Fig. 5A). No tumordeveloped in mice that received injection with BCG-823/Sur-Muttransfectants (Fig. 5B). All of these animals remained tumor free foran additional 2 months.

We next examined whether the decreased tumor growth rate wasassociated with increased apoptosis and/or decreased proliferation invivo. Consistent with the results in vitro, tumors derived from BCG-823/Sur-AS showed decreased expression of survivin protein com-pared with tumors derived from BCG-823/vector and parental cells(Fig. 5C). The tumors derived from BCG-823/Sur-AS had a higherapoptotic index (9 � 1%) than tumors derived from BCG-823/vectorand parental cells (4.0 � 1% and 3 � 1%, respectively; P � 0.01; Fig.6A). The proportion of Ki67-positive cells in tumors derived fromBCG-823/Sur-AS (8 � 1%) was significantly lower than that intumors from BCG-823/vector and parental cells (14 �1.0% and16 � 3%, respectively; P � 0.05; Fig. 6B). These results suggestedthat inhibition of tumor growth was attributable to increased sponta-neous apoptosis and decreased cell proliferation in vivo. We furtherexamined whether inhibition of survivin caused mitotic catastrophe invivo. Our results showed that tumors from BCG-823/Sur-AS cells5 Shuiping Tu, Xiaohua Jiang, Jian Ta Cui, and Benjamin C. Y. Wong, unpublished results.

Fig. 6. Inhibition of survivin induced apoptosisand mitotic catastrophe and suppressed cell prolif-eration in vivo. Tumors were subjected to TUNEL(A) and Ki67 (B) assays for respective determina-tion of apoptotic index (A, right panels; apoptoticcells are stained green) and proliferation index (B,right panels; small arrows indicate Ki67-positivecells). C, tumors were subjected to H&E stainingfor determination of the abnormal nuclei (rightpanels; small arrow indicates abnormally large andmultilobed nuclei cells). Each bar represents themean apoptotic index, proliferation index, or per-centage of abnormally large nuclei with 95% con-fidence intervals from two mice for 28 days afterinjections by two-tailed t test. Data represent themean � SE. n � 4. �, P � 0.01; #, P � 0.05(compared with control transfectants). Adjacentimages are representative fields (original magnifi-cation, �200) from tumors of the same size afterimplantation.

7730




exhibited a higher incidence of aberrant nuclei (7 � 2%) comparedwith tumors from parental cells and BCG-823/vector (2% and 3%,respectively; P � 0.01; Fig. 6C).

Stable Suppression of Survivin Reduced Tumor Angiogenesisin Vivo. To further investigate the effect of survivin inhibition onangiogenesis, we determined the expression of PECAM-1/CD31, awell-established endothelial cell marker (30), in tumors of the samesize derived from parental cells, BCG-823/vector, and BCG/Sur-AStransfectants. As shown in Fig. 7A, tumors derived from parental cellsand BCG-823/vector transfectants contained thick vessels predomi-nating over occasional discrete endothelial cells. In contrast, tumorsderived from BCG-823/Sur-AS transfectants contained mainly dis-crete endothelial cells, with occasional small and thin vessels. Tumor-associated neovascularization as indicated by MVD was quantified. Asignificant reduction of MVD in tumors derived from BCG-823/Sur-AS transfectant was found as compared with tumors derived fromBCG-823/vector transfectants and parental cells (P � 0.05; Fig. 7B).These results suggest that suppression of survivin reduces tumorangiogenesis in vivo.

DISCUSSION

In this study, we show that suppressing survivin expression orfunction causes spontaneous apoptosis and mitotic catastrophe in vivoin gastric cancer nude mice xenografts. Furthermore, our resultsindicate that suppression of survivin also inhibits de novo gastriccancer formation and angiogenesis in vivo.

Functional studies have demonstrated that suppression of survivinlevels in HeLa cells causes spindle defects and promotes apoptosis (1,11). It is conceivable that the suppression of survivin function ingastric cancer cells will also block the association of survivin tomitotic spindle, which will ultimately result in mitosis-initiated apo-ptosis. Gastric cancer in general is highly resistant to chemoradio-therapy and moderately resistant to apoptosis (31). A study has shownthat gastric cancer cells expressing a high level of survivin are morechemoresistant (25). This may be attributable to the effect of survivin

in facilitating gastric cancer cells to overcome both G0-G1 and G2-Mcheckpoint controls (25). These reports are consistent with our find-ings that suppression of survivin significantly increased the sensitivityof gastric cancer cells to both cytotoxic drugs and serum starvation.

Apart from apoptosis, we showed, for the first time, that stablesuppression of survivin by antisense and mutant (Cys84Ala) causedmitotic catastrophe in gastric cancer in vitro and in vivo. The molec-ular mechanism of mitotic catastrophe remains unclear, although ithas been shown that cell cycle checkpoint deficiencies, including theG1 checkpoint, G2 checkpoint, prophase checkpoint, and mitotic spin-dle checkpoint, promote mitotic catastrophe (13). Thus, antisensetargeting of survivin may result in dysregulation of mitotic spindlecheckpoint as well as defects in microtubule assembly and functionthat lead to mitotic catastrophe as suggested in other studies (18).Nevertheless, there are limited studies showing that transient expres-sion of Sur-Mut (Cys84Ala) causes aberrant cell mitosis (11). Aprevious study showed that synchronized cells expressing survivinThr34Ala mutant exhibited a relatively normal cell cycle (14). How-ever, a more recent study shows that stable expression of survivinThr34Ala mutant causes aberrant cytokinesis rather than apoptosis(32). Our results support the notion that stable suppression of survivincauses mitotic catastrophe, which leads to cell death. An increase inthe frequency of micronucleated tumor cells after radiotherapy andchemotherapy was suggested to be a positive prognostic marker oftreatment response. The mitotic catastrophe in vivo usually leads tonecrosis, thereby causing local inflammation that may be beneficial tothe antitumor effect (13, 33). In contrast, the process of apoptosis isnoninflammatory and therefore does not recruit the resources of theimmune system. With this consideration, inducing mitotic catastropheappears in general to be a desirable goal in cancer treatment.

We also demonstrated, for the first time, that suppression of sur-vivin in gastric cancer epithelial cells caused reduced development ofangiogenesis in association with tumor growth in vivo. The criticalrole of microvessel angiogenesis in tumor formation and tumor mi-gration has been widely recognized (34–35). In gastric carcinoma, the

Fig. 7. Inhibition of tumor angiogenesis in vivo. A, tumor angiogen-esis was assessed by immunohistochemical staining with PECAM-1/CD-31 on zinc fixative sections of tumors of the same volume derivedfrom parental BCG-823 cells, BCG-823/vector, and BCG-823/Sur-AStransfectants, respectively. Preparations were counterstained with hema-toxylin (original magnification �200). B, quantification of angiogene-sis, performed as described in “Materials and Methods.” The MVD wasthe average of the vessel counts obtained in the three sections. Areas ofhighest vascularization were chosen at low magnification (�100), andmicrovessel counting was performed at �200 on three chosen fields.Any immunoreactive endothelial cell or endothelial cell cluster, whichwas distinctly separate from adjacent microvessels, was considered as asingle countable vessel. Results were the mean of independent determi-nations by two investigators. Data represent the mean � SE. n � 4.P � 0.01.

7731




incidence of metastasis correlated with the number and density ofblood vessels (36, 37). Our study showed that blockade of survivinfunction inhibited angiogenesis in gastric cancer xenografts in nudemice. This is consistent with the previous reports that survivin playsa role in vascular endothelial growth factor-mediated endothelial cellprotection (19, 20, 38). Survivin up-regulation is one of the importantmechanisms for drug “resistance” in both tumor cells and tumor-associated blood vessel endothelial cells (39). It is suggested thatsuppression of survivin during angiogenesis removed the cytoprotec-tive effect of vascular endothelial growth factor, resulting in endothe-lial apoptosis and promoting rapid involution of three-dimensionalcapillary-like vessels in vitro (38). Our observations provide the firstin vivo evidence showing that, besides inducing apoptosis, targetingsurvivin exerts an additional antitumor effect by inhibition of thedevelopment of angiogenesis. Thus, targeting survivin expressiontherapeutically would not only compromise survival of the tumor cellbut could also indirectly affect activated endothelial cells of the tumorvasculature. In the present study, we showed that stable expression ofSur-Mut Cys84Ala completely prevented gastric cancer formationwith reduced development of angiogenesis in vivo.

In summary, our results showed that stable inhibition of survivinexpression and function resulted in spontaneous apoptosis and mitoticcatastrophe, enhanced sensitivity to cytotoxic drugs, and suppressionof de novo tumor formation with reduced development of angiogen-esis in gastric cancer xenografts in nude mice. Because of the pref-erential expression of survivin in gastric cancer but not in normaltissues (22, 23), these data suggest that targeting the survivin pathwayalone or with cytotoxic drugs may be useful in the treatment of gastriccancer.

REFERENCES

1. Li, F., Ambrosini, G., Chu, E. Y., Plescia, J., Tognin, S., Marchisio, P. C., and Altieri,D. C. Control of apoptosis and mitotic spindle checkpoint by survivin. Nature(Lond.), 396: 580–584, 1998.

2. Reed, J. C., and Bischoff, J. R. BIRinging chromosomes through cell division: andsurvivin’ the experience. Cell, 102: 545–548, 2000.

3. Giodini, A., Kallio, M. J., Wall, N. R., Gorbsky, G. J., Tognin, S., Marchisio, P. C.,Symons, M., and Altieri, D. C. Regulation of microtubule stability and mitoticprogression by survivin. Cancer Res., 62: 2462–2467, 2002.

4. Ambrosini, G., Adida, C., and Altieri, D. C. A novel anti-apoptosis gene, survivin,expressed in cancer and lymphoma. Nat. Med., 3: 917–921, 1997.

5. Monzo, M., Rosell, R., Felip, E., Astudillo, J., Sanchezm, J. J., Maestre, J., Martin,C., Font, A., Barnadas, A., and Avad, A. A novel anti-apoptosis gene: re-expressionof survivin messenger RNA as a prognosis marker in non-small-cell lung cancers.J. Clin. Oncol., 17: 2100–2104, 1999.

6. Swana, H. S., Grossman, D., Anthony, J. N., Weiss, R. M., and Altieri, D. C. Tumorcontent of the antiapoptosis molecule survivin and recurrence of bladder cancer.N. Engl. J. Med., 341: 452–453, 1999.

7. Adida, C., Haioun, C., Gaulard, P., Lepage, E., Morel, P., Briere, J., Dombret, H.,Reyes, F., Diebold, J., Gisselbrecht, C., Salles, G., Altieri, D. C., and Molina, T. J.Prognostic significance of survivin expression in diffuse large B-cell lymphomas.Blood, 96: 1921–1925, 2000.

8. Kato, J., Kuwabara, Y., Mitani, M., Shinoda, N., Sato, A., Toyama, T., Mitsui, A.,Nishiwaki, T., Moriyama, S., Kudo, J., and Fujii, Y. Expression of survivin inesophageal cancer: correlation with the prognosis and response to chemotherapy. Int.J. Cancer, 95: 92–95, 2001.

9. O’Connor, D. S., Grossman, D., Plescia, J., Li, F., Zhang, H., Villa, A., Tognin, S.,Marchisio, P. C., and Altieri, D. C. Regulation of apoptosis at cell division byp34cdc2 phosphorylation of survivin. Proc. Natl. Acad. Sci. USA, 97: 13103–13107,2000.

10. Chen, J., Wu, W., Tahir, S. K., Kroeger, P. E., Rosenberg, S. H., Cowsert, L. M.,Bennett, F., Krajewski, S., Krajewska, M., Welsh, K., Reed, J. C., and Ng, S. C.Down-regulation of survivin by antisense oligonucleotides increases apoptosis, in-hibits cytokinesis and anchorage-independent growth. Neoplasia, 2: 235–241, 2000.

11. Li, F., Ackermann, E. J., Bennett, C. F., Rothetmel, A. L., Plescia, J., Tognin, S.,Villa, A., Marchisio, P. C., and Altieri, D. C. Pleiotropic cell-division defects andapoptosis induced by interference with survivin function. Nat. Cell Biol., 1: 461–466,1999.

12. Olie, R. A., Simoes-Wust, A. P., Baumann, B., Leech, S. H., Fabbro, D., Stahel, R. A.,and Zangemeister-Wittke, U. A novel antisense oligonucleotide targeting survivin

expression induces apoptosis and sensitizes lung cancer cells to chemotherapy.Cancer Res., 60: 2805–2809, 2000.

13. Roninson, I. B., Broude, E. V., and Chang, B. D. If not apoptosis, then what?Treatment-induced senescence and mitotic catastrophe in tumor cells. Drug Resist.Updat., 4: 303–313, 2001.

14. Grossman, D., Kim, P. J., Schechner, J. S., and Altieri, D. C. Inhibition of melanomatumor growth in vivo by survivin targeting. Proc. Natl. Acad. Sci. USA, 98: 635–640,2001.

15. Mesri, M., Wall, N. R., Li, J., Kim, R. W., and Altieri, D. C. Cancer gene therapyusing a survivin mutant adenovirus. J. Clin. Investig., 108: 981–990, 2001.

16. Wall, N. R., O’Connor, D. S., Plescia, J., Pommier, Y., and Altieri, D. C. Suppressionof survivin phosphorylation on Thr34 by flavopiridol enhances tumor cell apoptosis.Cancer Res., 63: 230–235, 2003.

17. Kanwar, J. R., Shen, W. P., Kanwar, R. K., Berg, R. W., and Krissansen, G. W.Effects of survivin antagonists on growth of established tumors and b7-1 immunogenetherapy. J. Natl. Cancer. Inst. (Bethesda), 93: 1541–1552, 2001.

18. Altieri, D. C. Validating survivin as a cancer therapeutic target. Nat. Rev. Cancer, 3:46–54, 2003.

19. O’Connor, D. S., Daniel, S., Jeffrey, S., Schechner, S., Adida, C., Mesri, M.,Rothermel, A. L., Li, F. Z., Nath, A. K., Pober, J. S., and Altieri, D. C. Control ofapoptosis during angiogenesis by survivin expression in endothelial cells. Am. J.Pathol., 156: 393–398, 2000.

20. Tran, J., Rak, J., Sheehan, C., Saibil, S. D., LaCasse, E., Korneluk, R. G., and Kerbel,R. S. Marked induction of the IAP family antiapoptotic proteins survivin and XIAPby VEGF in vascular endothelial cells. Biochem. Biophys. Res. Commun., 264:781–788, 1999.

21. Papapetropoulos, A., Fulton, D., Mahboubi, K., Kalb, R. G., O’Connor, D. S., Li, F.,Altieri, D. C., and Sessa, W. C. Angiopoietin-1 inhibits endothelial cell apoptosis viathe Akt/survivin pathway. J. Biol. Chem., 275: 9102–9105, 2000.

22. Lu, C. D., Altieri, D. C., and Tanigawa, N. Expression of a novel antiapoptosis gene,survivin, correlated with tumor cell apoptosis and p53 accumulation in gastriccarcinomas. Cancer Res., 58: 1808–1812, 1998.

23. Okada, E., Murai, Y., Matsui, K., Isizawa, S., Cheng, C., Masuda, M., and Takano,Y. Survivin expression in tumor cell nuclei is predictive of a favorable prognosis ingastric cancer patients. Cancer Lett., 163: 109–116, 2001.

24. Krieg, A., Mahotka, C., Krieg, T., Grabsch, H., Muller, W., Takeno, S., Suschek,C. V., Heydthausen, M., Gabbert, H. E., and Gerharz, C. D. Expression of differentsurvivin variants in gastric carcinomas: first clues to a role of survivin-2B in tumourprogression. Br. J. Cancer, 86: 737–743, 2002.

25. Ikeguchi, M., Liu, J., and Kaibara, N. Expression of survivin mRNA and protein in gastriccancer cell line (MKN-45) during cisplatin treatment. Apoptosis, 7: 23–29, 2002.

26. Li, Q. F., Ou-Yang, G. L., Li, C. Y., and Hong, S. G. Effects of tachyplesin on themorphology and ultrastructure of human gastric carcinoma cell line BGC-823. WorldJ. Gastroenterol., 6: 676–680, 2000.

27. Chen, C., Liu, F-K., Qi, X-P., and Li, J-S. The study of chemiluminescence in gastricand colonic carcinoma cell lines treated by anti-tumor drugs. World J. Gastroenterol.,9: 242–245, 2003.

28. Ji, J. F., Chen, X., Leung, S. Y., Chi, J-T. A., Chu, K. M., Yuen, S. T., Li, R., Chan,A. S. Y., Li, J. Y., Dunphy, N., and So, S. Comprehensive analysis of the geneexpression profiles in human gastric cancer cell lines. Oncogene, 21: 6549–6556,2002.

29. Yokozaki, H. Molecular characteristics of eight gastric cancer cell lines established inJapan. Pathol. Int., 50: 767–777, 2000.

30. Giatromanolaki, A., Koukourakis, M. I., Theodossiou, D., Barbatis, K., O’Byrne, K.,Harris, A. L., and Gatter, K. C. Comparative evaluation of angiogenesis assessmentwith anti-factor-VIII and anti-CD31 immunostaining in non-small cell lung cancer.Clin. Cancer Res., 3: 2485–2492, 1997.

31. Shah, M. A. Recent developments in the treatment of gastric carcinoma. Curr. Oncol.Rep., 4: 193–201, 2002.

32. Temme, A., Rieger, M., Reber, F., Lindemann, D., Weigle, B., Diestelkoetter-Bachert, P., Ehninger, G., Tatsuka, M., Terada, Y., and Rieber, E. P. Localization,dynamics, and function of survivin revealed by expression of functional survivinD-sRed fusion proteins in the living cell. Mol. Biol. Cell, 14: 78–92, 2003.

33. Canman, C. E. Replication checkpoint: preventing mitotic catastrophe. Curr. Biol.,11: R121–R124, 2001.

34. Hanahan, D., and Folkman, J. Patterns and emerging mechanisms of the angiogenicswitch during tumorigenesis (review). Cell, 86: 353–364, 1996.

35. Fidler, I. J., and Ellis, L. M. The implications of angiogenesis for the biology andtherapy of cancer metastasis (minireview). Cell, 79: 185–188, 1994.

36. Maeda, K., Chung, Y. S., and Ogawa, Y. Prognostic value of vascular endothelialgrowth factor expression in gastric carcinoma. Cancer (Phila.), 77: 858–863, 1996.

37. Takahashi, Y., Cleary, K. R., Mai, M., Kitadai, Y., Bucana, C. D., and Ellis, L. M.Significance of vessel count and vascular endothelial growth factor and its receptor(KDR) in intestinal-type gastric cancer. Clin. Cancer Res., 2: 1679–1684, 1996.

38. Mesri, M., Morales-Ruiz, M., Ackermann, E. J., Bennett, C. F., Pober, J. S., Sessa,W. C., and Altieri, D. C. Suppression of vascular endothelial growth factor-mediatedendothelial cell protection by survivin targeting. Am. J. Pathol., 158: 1757–1765,2001.

39. Tran, J., Master, Z. B., Yu, J. L., Rak, J., Dumont, D. J., and Kerbel, R. S. A role forsurvivin in chemoresistance of endothelial cells mediated by VEGF. Proc. Natl. Acad.Sci. USA, 99: 4349–4354, 2002.

7732




2003;63:7724-7732. Cancer Res Shui Ping Tu, Xiao Hua Jiang, Marie C. M. Lin, et al. Tumorigenicity and Angiogenesis in Gastric Cancer

in VivoSuppression of Survivin Expression Inhibits

Updated version

http://cancerres.aacrjournals.org/content/63/22/7724

Access the most recent version of this article at:

Cited articles

http://cancerres.aacrjournals.org/content/63/22/7724.full.html#ref-list-1

This article cites 38 articles, 14 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/63/22/7724.full.html#related-urls

This article has been cited by 22 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

SubscriptionsReprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

[email protected] at

To request permission to re-use all or part of this article, contact the AACR Publications


http://cancerres.aacrjournals.org/content/63/22/7724

http://cancerres.aacrjournals.org/content/63/22/7724.full.html#ref-list-1

http://cancerres.aacrjournals.org/content/63/22/7724.full.html#related-urls

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

mailto:[email protected]


Documents

Suppression of survivin expression inhibits in vivo tumorigenicity and angiogenesis in gastric cancer