Loss of P53 and MCT-1 Overexpression Synergistically Promote Chromosome Instability and Tumorigenicity

Loss of p53 and MCT-1 Overexpression Synergistically Promote Chromosome Instability and Tumorigenicity

Ravi Kasiappan,1 Hung-Ju Shih,1,2 Kang-Lin Chu,1 Wei-Ti Chen,1 Hui-Ping Liu,3 Shiu-Feng Huang,1 Chik On Choy,1 Chung-Li Shu,1 Richard Din,4 Jan-Show Chu,5 and Hsin-Ling Hsu1

1Division of Molecular and Genomic Medicine, National Health Research Institutes, Zhunan, Miaoli County; 2Institute of Molecular Medicine, National TsingHua University, Hsinchu, Taiwan, Republic of China; 3Departments of Cardiovascular and Thoracic Surgery, Chang Gung Memorial Hospital, Taoyuan, Taiwan, Republic of China; 4School of Public Health, University of California at Berkeley, Berkeley, California; and 5Department of Pathology, Taipei Medical University, Taipei Medical University Hospital, Taipei, Taiwan, Republic of China

Abstract Introduction MCT-1 oncoprotein accelerates p53 degradation by means Cancer pathogenesis is involved in multiple pathways, in- of the ubiquitin-dependent proteolysis. Our present data cluding inactivation of tumor suppressors, activation of show that induction of MCT-1 increases chromosomal oncogenes, loss of cell differentiation, augmentation of prolifer- translocations and deregulated G2-M checkpoint in response ative activity, alteration of hormone receptor status, and increase to chemotherapeutic genotoxin. Remarkably, increases of metastatic potential (1). The fine balance between the in chromosome copy number, multinucleation, and proto-oncogene function and tumor suppressor activity plays a cytokinesis failure are also promoted while MCT-1 is induced critical role in regulation of cell growth, cell cycle, and genome in p53-deficient cells. In such a circumstance, the stabilization. Ras–mitogen-activated protein kinase/extracellular Genome aberrations are the hallmarkof tumorigenesis. signal-regulated kinase kinase–mitogen-activated protein Mitotic checkpoint controls the integrity of chromosome kinase signaling activity and the expression of metastatic structure (2-5). This checkpoint monitors chromosome align- molecules are amplified. Given a p53-silencing background, ment during metaphase and prevents premature progression MCT-1 malignantly transforms normal breast epithelial through mitosis. Specifically, it can stop mitotic cells from cells that are satisfactory for stimulating cell migration/ entering anaphase and prohibit chromosomes from moving adhesion and tumorigenesis. Detailed analyses of MCT-1 toward the spindle poles. Proper chromosome segregation is oncogenicity in H1299 p53-null lung cancer cells have mediated by centrosome cycle, mitotic spindle assembly, shown that ectopically expressed MCT-1 advances xenograft and mitotic kinase activation (6-8). Errors that are occurring tumorigenicity and angiogenesis, which cannot be at mitosis can result in mitotic catastrophes, characterized by completely suppressed by induction of p53. MCT-1 spindle collapse, multipolar spindles, or cytokinesis failure. counteracts mutually with p53 at transcriptional levels. Cells that have been encountered with mitotic catastrophes Clinical validations confirm that MCT-1 mRNA levels are can end up either entering apoptosis or exiting from mitosis differentially enriched in comparison between human lung with multinucleation or chromosome aberrations. Consequent- cancer and nontumorigenic tissues. The levels of p53 mRNA ly, the consequential chromosome instability comprises gains are comparatively reduced in a subset of cancer specimens, or losses of whole chromosome(s) (aneuploidy), transloca- which highly present MCT-1 mRNA. Our results indicate tions, amplifications/deletions, or breaks (9-11). that synergistic promotions of chromosomal imbalances Loss of p53 and hyperactivation of Ras–mitogen-activated and oncogenic potency as a result of MCT-1 expression protein kinase (MAPK) signaling are implicated in mitotic ab- and p53 loss play important roles in tumor development. normalities. A deficiency of p53 activity increases in the num- (Mol Cancer Res 2009;7(4):536–48) ber of centrosome along with mitotic defects, which leads to chromosome missegregation (12). In circumstances with little or no functional p53, cells are highly proliferative and nuclear structure rapidly disintegrates during aneuploidy development Received 9/9/08; revised 11/18/08; accepted 12/8/08; published online 4/16/09. Grant support: MG-097-PP-06 (H-L. Hsu) and National Science Council grant (13). Conversely, the p53-proficient cells can better escape 95-2320-B-400-012 (H-L. Hsu). from mitotic failures and limit the progress of chromosomal ab- The costs of publication of this article were defrayed in part by the payment of normality. Moreover, the mitotic abnormalities coupled with page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. centrosome amplification are induced by Ras–MAPK/extracel- Note: Supplementary data for this article are available at Molecular Cancer lular signal-regulated kinase (ERK) kinase (MEK)–MAPK ac- Research Online (http://mcr.aacrjournals.org/). Hsin-Ling Hsu, Division of Molecular and Genomic Medicine, National Health tivity, in which cells are greatly at riskof genomic instability Research Institutes, 35 Keyan Road, Zhunan, Miaoli County 350, Taiwan, (14). Several studies have indicated that activation of MAPK Republic of China. Phone: 886-37-246-166, ext. 35329; Fax: 886-37-586-459. signaling is important for cell transformation and genomic de- E-mail: [email protected] Copyright © 2009 American Association for Cancer Research. stabilization, which are induced by many oncoproteins, includ- doi:10.1158/1541-7786.MCR-08-0422 ing Mos, Ras, and Raf (15-17). Furthermore, activating MAPK

536 Mol Cancer Res 2009;7(4). April 2009 Downloaded from mcr.aacrjournals.org on September 24, 2021. © 2009 American Association for Cancer Research. MCT-1 Promotes Chromosome Aberration andTumorigenicity 537

(ERK1/2) and H-Ras selectively induces cell invasive pheno- (23). To further investigate the role of MCT-1 in response types by enhancing matrix metalloproteinase-9 (MMP-9) and to genotoxic stress, the chemotherapeutic agent ETO was cyclooxygenase-2 expression along with increasing invasion treated with MCT-1–expressing and vector control MCF- and migration abilities (18, 19). The chromosome instability 10A cells (Fig. 1A). In comparison with controls, the levels caused by p53 deficiency can be enhanced to a greater extent of phospho-Ser15 p53 and total p53 protein were relatively by the constitutively active MAPK pathway. This shows that a reduced by ectopically expressed MCT-1 either before ETO p53-dependent checkpoint mechanism can effectively prevent (0 hour) or after ETO treatment for 3 hours. In the latter case, the oncogene-mediated MAPK activation and chromosomal the cells were then recultured in an ETO-free medium for 24 instability. In addition, the Ras-MAPK pathway controls cell hours (3h→R24). Cytogenetic G-banding analysis subsequently cycle progression through G2 phase and affects the critical investigated chromosome deviations after ETO-treated cells re- checkpoint at G2-M transition following DNA damage (20). covered for 24 hours. Chromosome aberrations were quantified. Thus, oncogenic Ras activation and p53 dysfunction can dis- Table 1 indicated the frequencies of diverse and arbitrary rupt genomic integrity via a MAPK-dependent pathway. chromosomal abnormalities identified from each sample cohort The resistances to apoptosis and deregulation of DNA dam- (n = 30). Chromosomal exchanges, including reciprocal trans- age checkpoints have been identified in the oncoprotein MCT-1 locations (symmetrical) and dicentric formations (asymmetri- (multiple copies in a T-cell malignancy)–expressing cells cal), were predominantly increased under MCT-1 oncogenic (21-23). MCT-1 gene amplification was firstly recognized in stress (63%) compared with control group (30%). Chromosome the human B-cell and T-cell lymphoma (21). Ectopic expres- translocations were evidently increased >2-fold when MCT-1– sion of MCT-1 accelerates p53 protein degradation by activa- expressing cells underwent ETO damage, a topoisomerase II tion of the AKT-MDM2-proteosome pathway (23). Genome inhibitor that is widely used in cancer therapy. As well, chro- abnormalities, including chromosome amplification and trans- mosome deletions were more frequently observed in MCT-1– location, are accumulated as MCT-1 reduces p53 activity. expressing cells (20%) than controls (3.3%). Therefore, ge- MCT-1 is a CDC2/MAPK phospho-oncoprotein involved in nome surveillance machinery is impaired in MCT-1 oncogenic cell cycle regulation and DNA damage response (24, 25). stress. High frequency of chromosomal aberrations that are ad- Ectopically expressed MCT-1 activates the cell survival kinase, ditionally induced by MCT-1 could eventually enhance the tu- AKT, and prevents apoptosis induced by serum starvation (21). mor progression. Several lines of evidence show that MCT-1 implicates in the Consistent with MCT-1 induction diminishing p53 protein translational regulation and tumor development. MCT-1 phys- amounts (Fig. 1A), a quantitative reverse transcription-PCR ically associates with eukaryotic ribosomal complexes (26), (Q-RT-PCR) analysis indicated that MCT-1 significantly re- interacts with the cap complex through a RNA-binding motif, duced p53 mRNA production to 0.4-fold (Fig. 1B), revealing and recruits density-regulated protein (DENR/DRP) that con- that MCT-1 also down-regulated p53 at transcriptional stage. tains the SUI1 translation initiation domain (27). Oncogenic To inspect the oncogenic effect of MCT-1 in a p53-deficient MCT-1 can transform normal breast epithelial cells and en- background, p53 short hairpin RNA 2 was delivered into hance xenograft tumorigenicity of MCF-7 breast cancer cells MCF-10A cells. As examined by Q-RT-PCR, p53 gene ex- by enhancing invasiveness and decreasing apoptosis (22, 28). pressions were dramatically knocked down in both control However, MCT-1 tumorigenic effects underlying chromosome (control-p53) and MCT-1–expressing cells (MCT-1-p53; aberrations are unknown at present. Fig. 1B). Our present data provide direct clinical evidence that MCT-1 To evaluate the possible mechanism of MCT-1 affecting p53 mRNA levels are induced in human lung cancers. A p53- mRNA amounts, the drug that inhibits transcription (actinomy- silencing cellular background was used to recapitulate the cin D, 0.5 μg/mL) was added into culture and then endogenous MCT-1 oncogenic strength on promoting chromosome instabil- p53 mRNA decay was monitored for different time periods ity and mitotic abnormality. Cell migration and adhesion abil- (0-8 hours). Figure 1C data revealed that p53 mRNA turned ities are enhanced and tumorigenesis is established under a over more rapidly under MCT-1 oncogenic stress than that of circumstance of MCT-1 expressing/p53 silencing. Further, p53 mRNA that was comparatively stable in controls. This im- increasing MCT-1 levels in the p53-null lung cancer cells can plicates that MCT-1 activity has promoting effect on p53 advance xenograft tumorigenicity and angiogenesis but that mRNA turnover. cannot be completely suppressed by induction of p53. These results are the first to show that MCT-1 oncoprotein and p53 Ras Signaling Is Stimulated by MCT-1 Expression and reduction collectively enhance chromosome instability and p53 Deficiency tumorigenicity. As we examined cell proliferation kinase by Raf-1 RBD affinity assay (Fig. 2A), the amounts of active H-Ras were really elevated in both p53-deficient control (control-p53) and p53- Results deficient MCT-1–expressing cells (MCT-1-p53). After normal- Overexpression of MCT-1 Increases Etoposide-Induced izing to internal α-actin, the active H-Ras was determined to be Chromosomal Translocations a 2.2-fold increase in MCT-1-p53 group and a 1.36-fold enrich- In our previous findings, p53 and p21 protein levels are ment in control-p53 sample (Fig. 2B). On DNA damage (+ETO), decreased by MCT-1 overexpression in response to bleomy- H-Ras activity all significantly reduced but still remained at a cin (BLM) and etoposide (ETO) genotoxicity, and their re- 1.46-fold induction in MCT-1-p53 group in comparison with ductions are attenuated by the inhibition of MEK signaling control-p53 sample.

FIGURE 1. Constitutive expression of MCT-1 promotes chromosomal instability. A. MCF-10A cells were treatedwith ETO for 3 h andthen culturedin ETO-free medium for another 24h(3h→R24). The total p53 andphospho-p53 (Ser 15)proteins are reduced in MCT-1– expressing cells. B. Quantitative real-time PCR indicates that MCT-1 reduces p53 mRNA levels. The p53 short hairpin RNA 2 effectively inactivates p53 gene presentation in both control andMCT-1 –expressing cells. C. MCT-1expressionin- creases p53 mRNA decay as monitoredby mRNA turnover rate after actinomycin D blocking de novo transcription.

Signaling downstream of H-Ras, the phosphorylation of background (Fig. 3A). To our surprise, chromosome copy MEK1/2 and MAPK (ERK1/2) mitogenic cascade was enhanced numbers were increased notably as p53 silenced in MCT-1– in both MCT-1 and MCT-1-p53 cells even in response to genotox- expressing cells (MCT-1-p53). However, chromosome amplifi- in ETO (Fig. 2C). MEK1/2 phosphorylation in MCT-1-p53 cells cation did not show in p53-proficient MCF-10A (control) and moderately decreased, which was consistent with H-Ras deacti- was infrequently detected in p53-deficient controls (control- vation by ETO treatment (Fig. 2A and B), under which MAPK p53). Only increasing MCT-1 levels (MCT-1) also cannot dra- phosphorylation remained high. We speculate that H-Ras could matically alter chromosome numbers. In analysis of aneuploid be just one of upstream regulators of MEK1/2 and MAPK. and diploid populations after p53 loss, statistics (P < 0.001) MAPK response is followed on H-Ras-MEK signaling activity were acquired from at least 150 individual metaphase spreads and that its phospho-alteration could be temporally delayed. (Table 2). Comparatively, many more aneuploidies were de- In evaluation of cell proliferation status, equivalent numbers tected in MCT-1-p53 sample (42.8%) but less recognized in of the p53-silencing cells were cultured in the medium lacking control-p53 group (13.8%). sera and growth factors (−EGF/insulin) for 24 hours. Following To study if p53 induction can effectively prevent polyploidy incubation in epidermal growth factor (EGF)/insulin-containing generation, the p53-silencing MCF-10A was further restored medium (+EGF/insulin) for another 36 hours, the accumulative with p53 wild-type gene (Table 2). Surprisingly, ∼98.8% of cell numbers (Y axis) was indicated by the colorimetric 3-(4,5- p53-restored MCT-1-p53 cells (MCT-1-p53 + p53) exhibited an- dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) euploidy but no significant change of aneuploidy populations assay (Fig. 2D). In response to growth stimulation, a 1.8-fold in- (15.5%) in p53-restored control-p53 sample (control-p53 + crease of MCT-1-p53 cell populations was identified compared p53). Thus, p53 renovation cannot suppress but it actually ad- with control-p53 set. Therefore, p53 loss, along with MCT-1 in- vances aneuploidy; this could address the darkside of p53 in duction, profoundly induces H-Ras activity and mitogenesis. progressing chromosome instability under MCT-1-p53 situation. Flow cytometry analysis further confirmed that the polyploi- Additional Aneuploidy Is Occurring in MCT-1 Oncogenic dy frequencies were enhanced significantly in MCT-1-p53 cells Cells with p53 Silencing (57.2%) compared with control-p53 cells (3.9%; Fig. 3B). The Cytogenetic G-banding analysis was again to inspect the populations with DNA content more than 4N (polyploidy) were chromosomal abnormalities that happened in a p53-deficient even further increased when MCT-1-p53 cells (96.8%) were

Mol Cancer Res 2009;7(4). April 2009 Downloaded from mcr.aacrjournals.org on September 24, 2021. © 2009 American Association for Cancer Research. MCT-1 Promotes Chromosome Aberration andTumorigenicity 539

Table 1. ETO-Induced Chromosomal Aberrations

Events Ratio

Control cells Deletion 12 1 3.3% Breakage 2, 5, 5, 10, 17, 17, 21, 22, der(9), del(1) 10 33.3% Translocation t(2;?), t(6;10), t(7;11), dic(1;13), dic(2;4), dic(4;6), dic(der(9);X), dic(10;18), dic(X;16) 9 30% MCT-1–inducing cells Deletion 1, 5, 6, der(9), 18, 2 6 20% Breakage dup(1), i(8), 4, 2, 1, 7, 11, X 8 26% Translocation t(2;?), t(5;11), t(6;22), t(6;22), t(7;8), t(7;22), t(10;22), t(18;21), t(X;13), t(X;20), t(2;7), 19 63% dic(del(1);15), dic(5;6), dic(5;19), dic(8;14), dic(10;22), dic(2;12), dic(2;17), dic(2;X)

NOTE: In comparison with controls, MCT-1 overexpression greatly induces translocations and deletions. Abbreviations: t, translocation; del, deletion; dic, dicentric; dup, duplication; der, derivative. treated with the radiomimetic agent BLM for 24 hours. On served in BLM-untreated MCT-1–expressing cells, which indi- BLM treatment, however, 47.7% of control-p53 cells were cated that γ-H2AX formation before BLM treatment could be mainly arrested at G2-M phase and have no major increase in independent of MAPK activation. the polyploidy populations. Inhibition of MEK activity by To further confirm the effect of G2-M checkpoint in re- UO126 (20 μmol/L) during BLM treatment (BLM + UO126) sponse to BLM, ATM-CHK1-CDC25c signaling activation led to a moderate reduction of MCT-1-p53 polyploidy, 96.8% was examined (Fig. 3E). Compared with control-p53 sample, to 80.3%, indicating that UO126 modestly decreased 16.5% of the remarkable decrease in the phosphorylation of ATM polyploidy development. (Ser1981), CHK1 (Ser317), and CDC25c (Ser216) was quite In concordance with chromosomal instability, DNA dam- matched with G2-M checkpoint impairment identified in age–induced γ-H2AX foci that emerged before BLM treatment MCT-1-p53 cells. No significant alterations in CDC25c phos- (−BLM) were particularly found in the MCT-1-p53 group phorylation in BLM-treated or BLM-untreated cells could be (Fig. 3C). Along with promoting intrinsic γ-H2AX foci (non- due to loss of p53 inhibitory effect on CDC25c activation. BLM induced), the phospho-MAPK (p44/42) was greatly en- Cells cotreated with UO126 and BLM treatment (BLM + hanced (Fig. 3D), in which MEK-MAPK signaling coincided UO126) have only diminished phosphorylation on CHK1 with the progress of chromosomal instability in MCT-1-p53. but not ATM and CDC25c. Thus, MAPK hyperphosphoryla- However, no dramatic MAPK phosphorylation has been ob- tion and ATM signaling down-regulation happen together

FIGURE 2. Ras-MEK-MAPK signaling andcell proliferation are stimulatedby MCT-1 inducing/p53 silencing. A. H-Ras activity was analyzedby Raf-1 RBD affinity assay. Compared with p53 knockdown controls (control-p53), the active H-Ras is comparatively promotedas MCT-1–expressing cells abro- gate p53 expression (MCT-1- p53). B. On ETO treatment, H-Ras activity still increases in MCT-1-p53 cells. C. H-Ras downstream targets, phospho- MEK1/2 (pMEK1/2) andphospho-MAPK (pMAPK;ERK1/2), are also further activatedin MCT-1 andMCT-1-p53 cells. D. MTT assays indicate the accumulative number of cells (Y axis) after stimulation with EGF andinsulin.

FIGURE 3. Polyploidy is promoted by p53 silencing/MCT-1 expressing. A. G-banding analysis shows an increase of chromosomal copy numbers while p53 silencing andMCT-1 expressing. B. Flow cytometry data reveal that the polyploidy predominantly occurs in MCT-1-p53 cells either before (−BLM) or after BLM (+BLM) treatment. Adding MEK inhibitor throughout BLM treatment (+BLM + UO126) can moderately decrease polyploidy. C. Even lacking BLM, the intrinsic DNA damage γ-H2AX foci are increasedin MCT-1-p53 sample. D. In response to BLM, phospho-MAPK and γ-H2AX are stimulatedin MCT-1-p53 cells. E. Phosphorylation of ATM-CHK1-CDC25c is comparatively reduced in MCT-1-p53 cells that associate with deregulated G2-M checkpoint.

with polyploidy and γ-H2AX increments, particularly in a completed chromosome segregation followed on chromosomes MCT-1-p53 context. uniformly aligned on the metaphase plate (Fig. 4A). However, the cellular chromosomes of p53-silencing MCT-1 cells (MCT- More Mitotic and Nuclear Abnormalities Happen as MCT- 1-p53) aligned at the spindle equator but kept oscillating with 1–Expressing Cells Lose p53 low amplitude around the metaphase plate (Fig. 4B). Desegre- In examination of nuclear morphology and mitotic process, gated chromosome bridges were observed frequently during the majority of p53 knockdown controls (control-p53) have anaphase. In analysis of 250 mitotic cells, the percentages of

Table 2. Aneuploidy Frequencies Are Increased in MCT-1-p53 and MCT-1-p53+p53 Conditions

Control-p53 MCT-1-p53 Control-p53 + p53 MCT-1-p53 + p53

Diploid (%) Aneuploid (%) Diploid (%) Aneuploid (%) Diploid (%) Aneuploid (%) Diploid (%) Aneuploid (%)

Samples A861458428416099 B901062388614099 C901052488218298 D821860408614099 E83175446NDNDNDND Mean ± SD 86.2 ± 3.4 13.8 ± 3.4 57.2 ± 3.7 42.8 ± 3.7 84.5 ± 1.5 15.5 ± 1.5 1.25 ± 0.4 98.8 ± 0.4

NOTE: From independent experiments, the frequencies of diploidy and aneuploidy were evaluated between p53-deficient control (control-p53), p53-deficient MCT-1 (MCT-1-p53), and p53-deficient cells that reconstituted with p53 gene (control-p53 + p53 and MCT-1-p53 + p53). Abbreviation: ND, not determined.

lagging chromosomes in anaphase (mitotic abnormalities) were To more characterize nuclear activity of the multinucleate more detected in MCT-1-p53 (30.97%) than those identified in cells, the nuclear mitotic apparatus protein (NuMA) was coim- control-p53 (13.08%; Fig. 4C). Further probing nuclear integ- munostained with the endogenous MCT-1 (Fig. 4F). The numb- rity in telophase (Fig. 4D), 17.39% of MCT-1-p53 cells exhib- ers of multinucleate MCT-1-p53 cells (top) were greatly ited binuclei, trinuclei, or micronuclei (Fig. 4E). Conversely, increased over control-p53 sample (bottom). We found three these dramatic nuclear aberrations were barely detected in interphase nuclei (arrow), a mitotic nucleus with chromosome p53-deficient controls (2.34%). Therefore, abnormal cytokine- alignment at the metaphase plate (star), and some micronuclei sis can cause intrinsic alterations of chromosome copy numbers (asterisk) that were all enclosed within a single cell. This in MCT-1-p53 cells. proves that unsynchronized cell cycle progression at different

FIGURE 4. Mitotic aberrations happen in MCT-1-p53 cellular context. A. The p53-deficient control cells have intact nuclear structure and regular mitosis progression. Scale bars, 15 μm. B. MCT-1-p53 cells exhibit lagging chromosomes in anaphase. C. The mitotic abnormalities are comparedbetween control- p53 andMCT-1-p53. D. At telophase, MCT-1-p53 cells show cytokinesis failure, leading to multinucleation or nuclear disintegration. Scale bars, 10 μm. E. The numbers of cells with cytokinesis incompletion and nuclear aberrations are quantified. Columns, average of three independent experiments; bars, SD. F. NuMA immunostaining shows irregular nuclear morphology of the multinucleatedMCT-1-p53 cells. Unsynchronizedcell cycle is progressing in a multi nuclear cell (denoted by arrow, asterisk, and star). Scale bars, 15 μm.

FIGURE 5. Metastatic potential andtumorigenesis inducedby p53 silencing/MCT-1 expressing. HIF-1 α mRNA (A), MMP-9 mRNA (B), andintegrin β4 mRNA levels (C) are highly presentedin MCT-1 andMCT-1-p53 cells comparedwith control andcontrol-p53 cells. In concordancewith promoting metastatic potential, MCT-1 andMCT-1-p53 cells adheringto laminin-coatedplate ( D) andmigrating through polyethylene terephthalate –coatedmembrane ( E) increase more than control andcontrol-p53 cells do. F. In comparison with normal lung tissue (set as 1), 10 of 11 (90.9%) human lung tumor specimens show evidence of MCT-1 mRNA increments as determined by Q-RT-PCR. MCT-1 mRNA is rarely detected in the nontumorigenic tissues. G. The levels of p53 mRNA in lung tumors were comparedwith the relative normal tissues (set as 1). Low p53 mRNA levels are detectedina tumor sample ( 11), which highly expresses MCT-1. MCT-1 is undetectable in the tumor with p53 abundance (1). Columns, mean from three different assays; bars, SD.

nuclear compartments of a multinuclear cell perpetuates the rate Thus, stimulations in HIF-1α,MMP-9,andintegrinβ gene of aneuploidy. activation are p53 independent. When cell adhesion ability in a serum-free condition was analyzed with a laminin-coated MCT-1 Induction and p53 Silence Collectively Promote 96-well plate, MCT-1 and MCT-1-p53 cells showed a dramatic Tumorigenesis adhesion activity compared with control or control-p53 cells The metastatic potential was examined in nontumorigenic (Fig. 5D). Moreover, by using 5% horse serum as a chemoat- MCF-10A (Fig. 5A-C). A 1.96-fold increase in hypoxia-induc- tractant, MCT-1 and MCT-1-p53 cells revealed high migratory ible factor-1α (HIF-1α) mRNA was identified in MCT-1-p53 ability through a Boyden chamber (Fig. 5E). The elevation of cells (Fig. 5A). Moreover, MMP-9 mRNA was found to be a HIF-1α, MMP-9, and integrin-β4 mRNAs has implicated that 1.99-fold enhancement in MCT-1-p53 circumstance (Fig. 5B). MCT-1 up-regulates several genes that promote cell survival, MCT-1-p53 cells also presented a 1.7-fold amplification in in- angiogenesis, migration, invasion, and adhesion to extracellular tegrin β4 mRNA relative to control-p53 sample (Fig. 5C). In matrix proteins (29-31). fact, oncogenic MCT-1 itself has already greatly elevated the To examine if putting additional MCT-1 oncogenic effects transcripts of HIF-1α (1.74-fold), MMP-9 (1.43-fold), and in- into the p53-silencing condition are sufficient for tumorigen- tegrin β4 (1.93-fold) in comparison with comparative controls. esis, the p53-silencing MCF-10A cells at cultivated p27 and

p38 were s.c. injected into nude mice. Symptoms of tumor tumor weights, MCT-1/H1299 tumors have a 10-fold incre- development were only found in 25% of xenografts, which ment in hemoglobin quantities than control/H1299 tumors. were injected with the late passage of MCT-1-p53 cells Angiogenic effect was further characterized by CD31 immu- (p38; Table 3). On the contrary, no sign of tumor develop- nohistochemical analysis that corresponded to microvessel ment was detected in the mice inoculated with MCT-1–ex- density. The data show that MCT-1/H1299 tumors have more pressing or p53-silencing control cells. This shows that than 2.5 times increase in vascular counts relative to those genomic and mitotic aberrations that accumulated in MCT- control/H1299 tumors. 1 overexpression/p53 loss are enough for malignancy and MCT-1 mRNA levels were further evaluated among nor- tumorigenesis. mal and tumor tissues from nude mice by RT-PCR analysis To obtain direct clinical evidence, 11 pairs of non–small cell (Supplementary Fig. S1). The internal 18S rRNA levels were lung cancer tissue along with adjacent nonneoplastic lung tis- constitutively expressed among samples (denoted with aster- sues were examined. Among them, four samples were squa- isks). MCT-1 transcripts (indicated with arrows) were partic- mous cell carcinoma (4, 5, 10, 11) and eight specimens were ularly elevated in MCT-1/H1299 tumors (M1 and M2) but adenocarcinoma (1-3, 6-9), respectively. All of them had done undetectable in normal tissues and control/H1299 tumors EGF receptor (EGFR) mutation analysis, and four of the eight (C1 and C2). Detailed quantitative analysis indicated that adenocarcinoma tumor tissues had EGFR mutations (1, 6, 8, the xenograft tumors that emerged from MCT-1/H1299 pre- 12). Q-RT-PCR analysis indicated that MCT-1 gene was differ- sented ∼20-fold MCT-1 mRNA levels in comparison with entially induced in nearly all human lung cancer samples control/H1299 xenograft tumors (Fig. 6C). Consistent with (90.9%) in comparison with relative nontumorigenic lung tis- angiogenesis stimulation, both H-Ras and HIF-1α mRNA le- sues that expressed extremely low levels of MCT-1 (Fig. 5F). vels were significantly enriched in MCT-1/H1299 xenograft In an attempt to find the correspondence between MCT-1 and tumors as assessed with Q-RT-PCR analysis (Supplementary p53 presentations, p53 mRNA levels were analyzed in lung Fig. S2). These also correspond with the enhanced H-Ras sig- cancer samples in parallel with their comparative normal tissues naling, proliferation, and migration/adhesion ability observed (Fig. 5G). MCT-1 mRNA presentation was greatly amplified in in MCF-10A cells with MCT-1 expressing/p53 silencing a lung tumor that has comparatively lower amounts of p53 (Figs. 2 and 5D and E). mRNA (11). In contrast with that, two of cancer samples (1, 5) presented low MCT-1 mRNA levels but have high Knock-in p53 Gene Cannot Completely Suppress MCT-1 amount of p53 transcripts. These clinical data reveal that Tumorigenic Effects MCT-1 status could play an important role in the etiology of To examine if the exogenous p53 can functionally compro- human lung cancer. mise MCT-1 tumorigenic activity, H1299 cells that were pri- marily transfected with pLXSN vector or MCT-1 oncogene MCT-1 Induction Promotes Tumor Growth in a p53-Null were further introduced with pLHCX vector or wild-type p53 Background gene. Similar to Fig. 1C data, p53 mRNA was in a 0.4-fold To recapitulate whether MCT-1 tumorigenic outcomes can reduction in MCT-1 + p53 cells compared with control + p53 be augmented in p53-null lung cancer cells, the vector control sample (Fig. 6D). – H1299 (control/H1299) and MCT-1 expressing H1299 Subsequently, the p53-reconstituted H1299 cells (MCT-1 + (MCT-1/H1299) were s.c. injected into nude mice. Conse- p53 and control + p53) were s.c. injected into nude mice. quently, larger tumors with more neovascularization emerged Unexpectedly, the ectopic MCT-1 + p53 H1299-injected mice in MCT-1/H1299 xenografts than those in control/H1299 xe- still developed larger tumors (4.3-fold increase) contiguous nograft tumors (Fig. 6A). A 20-fold increase in tumor weights with more hemoglobin amounts (2.3-fold increase) than those was found in MCT-1/H1299 xenografts compared with con- evolving from the control + p53 cells (Table 5). In compar- trol/H1299 xenografts. The incidence of tumors that devel- ison with Table 4 data, p53 restoration certainly reduced the oped from MCT-1/H1299 was 4-fold higher than from extent of tumor burdens and hemoglobin amounts caused by control/H1299 (Table 4). Hemoglobin amounts in these tu- MCT-1/H1299. As reported by Q-RT-PCR study in H1299 mors were measured by the QuantiChrom hemoglobin assay. tumors, MCT-1 mRNA expression was a 0.4-fold decrease Compared with the hemoglobin standard and normalized with when p53 was renovated (compared between MCT-1 and MCT-1 + p53 xenografts; Fig. 6F). As well, H1299 cells that ectopically coexpressed MCT-1 and p53 (MCT-1 + p53) have more than a 0.43-fold decrement in MCT-1 mRNA amounts Table 3. Tumorigenesis Is Caused by MCT-1 Overexpression Along with p53 Loss than MCT-1/H1299 (Fig. 6G). These data suggest that p53 might act as a transcriptional factor that negatively regulates Cell Types Tumor Incidence (%) Tumor Weight (g) MCT-1 gene expression. Even if p53 induction is capable of suppressing MCT-1 mRNA production, no complete regres- Control-p53 (p27) 0 ND sion in tumorigenicity was found. Therefore, MCT-1 induc- MCT-1-p53 (p27) 0 ND Control-p53 (p38) 0 0 tion enhancing H1299 tumorigenicity is not simply because MCT-1-p53 (p38) 25 0.36 of p53 deletion, but it also synergizes with other p53-inde- Total injection times, 2.5 mo pendent oncogenic factors. In conclusion, MCT-1 overexpres- NOTE: The p53-silencing MCF-10A cells premixed with Matrigel were injected sion reinforces tumorigenicity and metastaticity while p53 is into nude mice for 45 d. Only 25% of MCT-1-p53 xenografts emerge tumors. missing.

FIGURE 6. MCT-1 promotes tumor development in a p53-null background. A. The tumor incidences and burdens are significantly enhanced in MCT-1/ H1299 xenografts comparedwith control/H1299 xenografts. B. MCT-1 overexpression stimulates the angiogenesis pathway because of increase in hemoglobin amounts andthe endothelialmarker CD31 highlight microvessels ( arrows). C. MCT-1 transcripts are qualitatively enhancedin MCT-1/H1299 tumors. Columns, mean; bars, SD from three independent assays. D. MCT-1 still diminishes p53 mRNA production after p53 gene transferring into H1299. E. Nevertheless, p53 induction cannot fully repress tumorigenicity and hemoglobin amounts caused by MCT-1-p53. MCT-1 mRNA levels are found to be reduced by p53 restoration either in H1299 xenografts (control, MCT-1, control + p53, and MCT-1 + p53; F) or in H1299 cells (G).

Discussion chromosomal abnormalities (22, 23). We speculate that the Our previous workhas identified that overexpression oncogenic chromosome aberrations and tumorigenicity will of MCT-1 significantly reduces p53 function, deregulates enormously amplify while putting MCT-1 oncogenic stress DNA damage responses, and increases the incidence of in already p53-null or p53-deficient cells, which cannot

Table 4. Tumor Development and Angiogenesis Are (Table 3). MCT-1 overexpression itself may be not only di- Promoted by Increasing MCT-1 in a p53 Null rectly oncogenic but could also induce a constellation of invasive and metastatic factors (e.g., HIF-1α, MMP-9, and H1299 Tumor Tumor Hemoglobin Capillary (p53 Null) Incidence (%) Weights (g) (mg/g Tissue) Density integrin β4) by its transcriptional or translational regulation Xenograft function. The enhancing H-Ras gene expression followed by an extremely active Ras-MEK-MAPK signaling cascade Control 20 (1) 0.06 (1) 0.33 (1) 1.7 (1) MCT-1 80 (4) 1.29 (21.5) 3.43 (10.4) 4.3 (2.5) reinforces tumorigenic outcomes and angiogenic effects (18, 19, 32). Oncogenic MCT-1 participates in stimulating the NOTE: MCT-1 overexpression stimulates the angiogenesis pathway because Ras-MEK-MAPK cascade (Fig. 2A-C) as well activates increase in hemoglobin amounts and the endothelial marker CD31 highlight HIF-1α, MMP-9, and integrin β4 gene expression microvessels. (Fig. 5A-C), which are overexpressed in diverse malignancies. These also speakto MCT-1 –promoting abilities in the properly maintain genomic integrity and monitor mitotic cell migration and invasion (Fig. 5D and E). checkpoint. Our current data show that MCT-1 counteracts Ras/Raf/MAPK signaling cascade has a biochemical link p53 function and further promotes chromosomal transloca- with p53 transcription and p53 activity. Ras inhibition increases tions, deletions, and amplifications in the absence of p53 p53 mRNA levels, indicative of a Ras-dependent mechanism (Figs. 1 and 3). that regulates p53 transcriptional activation (33). Because The hyperactive Ras-MEK-MAPK signaling cascade is re- the MEK inhibitor (UO126) can partly attenuate p53 and p21 related to centrosome amplification, multipolar spindles, and ductions (23), this suggests that MCT-1 could deregulate p53 chromosome bridges (14). Silencing of p53 in MCT-1–induc- transcription by constitutively activating Ras-Raf-MAPK. ing cells (MCT-1-p53) can stimulate Ras-MEK-MAPK activ- Our data indicate that by escalating MCT-1 levels in a p53-null ity (Fig. 2), explaining that cell proliferation is less reliant background (H1299) that carried chromosomal inequality and on sera and growth factors. We are exploring every possibility oncogenic mutations, the tumorigenicity of MCT-1 is manifestly of suppressing the progress of polyploidization induced reinforced. As tumor growth and angiogenesis are synergistical- by oncogenic MCT-1 and p53 deficiency. The polyploidy in- ly promoted in MCT-1/H1299 xenografts (Fig. 6A; Table 4), cidences are reduced moderately by a MEK inhibitor, UO126 these could relate to stimulation in H-Ras and HIF-1α (Supple- (Fig. 3B), revealing that the MEK signaling is rationally mentary Fig. S2) and provide in vivo confirmation for MCT-1 linkto chromosomal aberrations that have takenplace in a oncogenicity predominantly in a p53-null background. Our data MCT-1-p53 condition. The deregulated cell growth and mito- also show that by means of p53 gene transferring cannot totally genesis could produce not only multinuclei but also micronu- repress MCT-1/H1299 (p53 null) tumorigenicity (compare Fig. clei. Micronucleation can further cause the cells to lose large 6B with Fig. 6E; Tables 4 and 5). Previous studies have indicated segments or to generate small fragments of chromosomes that MCT-1–expressing MCF-7 (p53 proficient) also can pro- via breakage of the anaphase bridge. Circumstances of the in- mote angiogenesis in xenografting tumorigenicity (28). Reduc- duction of γ-H2AX foci that represented several intrinsic tion of MCT-1 mRNA is identified in p53-reconstituted H1299 DNA breaks are incompletely repaired in a MCT-1-p53 con- cells and xenograft tumors (Fig. 6F and G), implying that p53 text before encountering with the genotoxin (Fig. 3D and E). could antagonize mutually with MCT-1 at a transcriptional level. Cytokinesis failure also contributes to chromosomal missegre- Reciprocally, MCT-1 overexpression works against p53 function gation and increases the riskof micronuclei, giant nuclei, or and promotes advanced tumorigenicity in the absence of p53, multinuclei (Fig. 4B-F), by which MCT-1 induction further besides that MCT-1 has other tumorigenic functions that are increases aneuploidy in a p53 knockdown background p53 independent. (Fig. 3A and B; Table 2). Furthermore, unsynchronized cell A very important finding is that increasing MCT-1 tran- cycle progression in a multinucleate MCT-1-p53 cell can be scripts are identified in the majority of human lung cancer spe- another major cause of chromosome inequality (Fig. 4F). In cimens examined at this point (Fig. 5F), implicating that MCT- summary, MCT-1 confers its oncogenic influences by affect- 1 hyperactivation could associate with the cancer susceptibility. ing many signaling pathways, which collectively enhance ge- The clinical evidence might disclose the counteracting relation nomic mutations after p53 loss. Significant oncogenic between MCT-1 and p53, in which p53 mRNA levels are com- outcomes are not from a single cause but rather from a conflu- paratively low in a sample that highly expresses MCT-1 ence of factors provoked by MCT-1 oncogenicity and p53 absence. The chromosomal instabilities identified in a MCT-1- p53 situation are the collective consequences of deregulation Table 5. Restoration of p53 Cannot Suppress MCT-1 of DNA damage checkpoints and mitotic aberrations. The fun- Tumorigenic Effects damental mechanism by which MCT-1 concomitantly exacerbates mitotic and nuclear aberrations is under investigation. H1299 Tumor Hemoglobin Capillary – (p53 Add-Back) Weights (g) (mg/g Tissue) Density Given a MCT-1 inducing/p53-eradicating circumstance, Xenograft nontumorigenic MCF-10A cells are malignantly transformed and provide evidence for tumor development in the xenograft Control + p53 0.44 (1) 0.18 (1) 2.7 (1) nude mice. All these in vitro and in vivo findings show for MCT-1 + p53 1.88 (4.3) 0.41 (2.3) 5.4 (2.0) the first time that MCT-1 induction, along with p53 deficien- NOTE: p53 induction cannot fully repress tumorigenicity and hemoglobin cy, is sufficient for provoking tumorigenesis in the mammals amounts caused by MCT-1.

(Fig. 5G, 11). Vice versa, MCT-1 is barely detectable in the tu- RPMI 1640 at both s.c. sites. The p53-silencing MCF-10A cells mor sample that is p53 abundant (Fig. 5G, 1). (2 × 106) were premixed with equal volume of ice-cold BD Our current findings also open some intriguing questions, in- Matrigel matrix (10 mg/mL; BD Biosciences) before injecting cluding how MCT-1 can induce tumorigenesis in a p53-null or into nude mice. When tumor size had reached approximately in a p53-preserve backgrounds, how does MCT-1 stimulate me- 4 to 6 mm, tumors were resected, weighted, and processed tastatic potential after p53 deficiency, whether oncogenic MCT- for immunohistochemistry or Q-RT-PCR analysis. 1 and p53 loss synergistically up-regulate the metastatic mole- Hemoglobin amounts in the tumors were measured by spec- cules via a posttranscriptional mechanism as recently reported trophotometry using the QuantiChrom hemoglobin assay kit (34), and whether oncogenic MCT-1 turns tumor suppressor (BioAssay Systems). Tumors excised from mice were immedi- p53 into the darkside of contributing to tumorigenesis (35). Ad- ately washed with PBS, immersed in 1 mL sterile distilled vance knowledge in MCT-1 function could illuminate the fun- water, and incubated on an orbital shaker for 10 min at room damental mechanism of tumorigenesis and help the design of temperature. Following centrifugation at 14,000 rpm for therapeutic target. 15 min, 50 μL of the supernatant were incubated with 200 μL of the hemoglobin assay reagent for 5 min at room temperature. Hemoglobin concentrations were calculated based on Materials and Methods absorbance absorption at 400 nm. Compared with the standard, Antibodies and Reagents hemoglobin quantities were indicated as absorbance sample/ Antibodies recognizing p53 and actin were acquired from absorbance standard = mg/g wet tissue. Santa Cruz Biotechnology; phospho-CHK1 (Ser317), phospho-CDC25c (Ser216), phospho-p53 (Ser15), phospho-MAPK Human Lung Cancer Tissues and Quantitative Real-time (p42/p44), and ERK1/2 were from Cell Signaling Technology; RT-PCR Analysis phospho-ATM (Ser1981), γ-H2AX, and H-Ras were from Up- Eleven pairs of surgically resected fresh frozen non–small state Biotechnology; α-actin or β-actin was from Abcam; and cell lung cancer tissues along with adjacent nonneoplastic lung V5 epitope antibody was from Invitrogen. BLM, ETO, and ac- tissues were obtained from the tissue bankof Chang-Gung tinomycin D were acquired from Sigma and UO126 was from Memorial Hospital (Taoyuan, Taiwan), which were approved Cell Signaling Technology. by the Institutional Review Board of Chang-Gung Memorial Hospital. MCT-1 Overexpression and p53 Knockdown Total RNA was extracted from the tissues using Trizol re- MCF-10A cells were transfected with pLXSN MCT-1-V5 agent (Invitrogen). cDNA was synthesized from 2 μgoftotal using viral supernatants collected from PT67 cells and subse- RNA using an oligo(dT)12-18 primer and SuperScript II reverse quently transfected with pMKO.1 puro p53 short hairpin transcriptase (Invitrogen). The Q-RT-PCR was done with RNA 2 (Addgene) using Lipofectamine 2000 (Invitrogen). nucleotide61MCT-1sense(5′-GAGCGGAAGTAGTCA- Four cohorts of transfectants (mass culture) were established, GATTT-3′) and nucleotide 431 MCT-1 antisense (5′- – including pLXSN-control (control), MCT-1 overexpressing TGTTCATGGCATCGGACTAT-3′) primers. Reaction cells (MCT-1), p53-silencing control (control-p53), and mixtures (20 μL) contained 150 ng cDNA, 2 μmol/L primers, – MCT-1 expressing/p53-silencing cells (MCT-1-p53). These and 1× SYBR Green Master Mix (Applied Biosystems). Reac- cellswereculturedinDMEM/F12mediumcontaining5% tions were run on the ABI Prism 7900 Fast Real-Time PCR μ μ horse serum, 20 ng/mL EGF, 10 g/mL insulin, 0.5 g/mL hy- System in triplicate. The reaction was conducted as follows: drocortisone, 100 ng/mL cholera toxin, 100 units/mL penicillin, 95°C for 10 min followed by 45 cycles of a 15-s denaturing μ μ 100 g/mL streptomycin, and selection antibiotic (100 g/mL at 95°C and 1-min annealing at 60°C. The RT-PCR products μ G418 or 0.5 g/mL puromycin). were resolved by 2% agarose gel electrophoresis and stained Raf-1 RBD Affinity Assay for Active H-Ras with ethidium bromide. The mRNA levels were calculated. Cy- ΔC C − C Exponentially growing MCF-10A cells were cultured with cle threshold ( t)= t target gene (MCT-1) t endogenous α 40 μmol/LETOfor1h.ActiveRaswasisolatedonRaf-1 control (18S rRNA gene). H-Ras, HIF-1 , MMP-9, integrin β RBD-glutathione agarose according to the manufacturer's in- 4, and p53 specific primers were respectively designed by structions (Upstate Biotechnology). Cells were extracted with Primer Express software to ensure a single 69-, 76-, 54-, 114-, magnesium-containing lysis buffer and precleared with gluta- and 72-bp amplicon. The probes were labeled with NFQ thione beads. Raf-1 RBD agarose (5-10 μg of protein) was in- (quencher) and FAM (reporter) and synthesized by Integrated cubated with 500 μg to 1 mg of lysates and gently rotated at 4° DNA Technologies (Applied Biosystems). The standard Taq- β C for 30 min. The beads were washed thrice with magnesium- man assays were analyzed after normalizing to -actin. containing lysis buffer. Raf-1 RBD-associated H-Ras was resolved by SDS-PAGE and immunoblotting. Cell Proliferation MTT Assay MCF-10A cells were subcultured in the basal DMEM/F12 Tumor Xenografts and Hemoglobin Assay medium without growth factors and horse sera. Fifty microliters Non–small cell lung cancer cells (H1299) were overex- of cell suspension (1 × 105 cells/mL) were seeded into each 96- pressed in MCT-1 and cultured as previously described (23). well plate and incubated at 37°C for 24 h. After starvation, cells Eight-week-old female BALB/c nude mice were injected with were supplemented with 50 μL DMEM/F12 medium contain- MCT-1–expressing H1299 or the vector control H1299. Each ing growth factors (40 ng/mL EGF and 20 μg/mL insulin) for mouse was injected with 2 × 106 cells suspended in 100 μL 36 h. Cell Proliferation Kit I (Roche) was used as follows: cells

were incubated with 10 μL MTT labeling reagent for 4 h and plates. The preequilibrated medium (800 μL) containing 5% then incubated with 100 μL solubilization solution for 24 h at horse serum was used as a chemoattractant. MCF-10A cells 37°C. The purple formazan crystals were analyzed at an absor- (5 × 104) were suspended in 100 μL of serum-free medium bance of 595 nm using a 96-well plate spectrophotometer and added into polyethylene terephthalate–coated insert and (SpectroMAX Plus, Molecular Devices). then incubated in 5% CO2 at 37°C for 1 h. The unattached and unmigrated cells were rinsed off and gently wiped off with Immunofluorescence Microscopy a cotton swab. The membranes with migratory cells were fixed Cells were fixed with 3.5% formaldehyde in PBS for in 70% methanol for 30 min, washed with PBS, and stained 15 min at room temperature and then permeabilized with with 10× Mayer's hematoxylin (DakoCytomation). Six ran- − ice-cold acetone for 3 min at 20°C. The samples were incu- domly selected fields per membrane were analyzed. The experi- bated with primary antibody for 2 h followed by washing ments were conducted in triplicate. with PBS. Alexa Fluor 488–coupled goat anti-mouse or Alexa Fluor 543–conjugated goat anti-rabbit secondary antibodies Cytogenetic Analysis (Invitrogen, Molecular Probes) were incubated and counter- MCF-10A cells were treated with ETO (10 μmol/L) for stained with 4′,6-diamidino-2-phenylindole for 1 h in the 3 h and then cultured in ETO-free medium for 24 h. The cells dark. Images were analyzed by a Nikon Optiphot-2 upright were arrested with colcemid (0.1 μg/mL; Invitrogen) for 4 h. fluorescence microscope at a 100× objective or a Leica Both detached and adherent cells were harvested and gently TCS NT confocal microscope at a 63× objective. Data shown suspended in hypotonic solution (0.075 mol/L KCl) at 37°C were representative of three independent experiments. for 8 to 10 min. Subsequently, five drops of Carnoy's fixative (methanol and glacial acetic acid at 3:1 ratio) were added and Genotoxin Treatment, Immunoblotting, and Flow Cytometry then centrifuged at 1,200 rpm for 5 min. Following fixation for Analysis three times and kept at 4°C for at least 1 h, cells were sus- Genotoxic agents (6 milliunits BLM or 40 μmol/L ETO) pended in fresh fixative solution. Metaphase spreads were were used to induce double-strand DNA breaks. Whole-cell dropped onto ice-cold wet slides and dried at 100°C for extracts were prepared using CytoBuster protein extraction 25 min. G-banding analysis was conducted with 0.04% trypsin reagent (Novagen) by incubating on ice for 15 min. Extracts for 45 s and stained for 1 min in 0.075% Wright's eosin meth- were cleared by centrifugation at 4°C for 15 min. Protein sam- ylene blue diluted with Hank's buffer. Slides were washed, ples (60 μg) were heat denatured with NuPAGE lithium dode- mounted, and photographed. cyl sulfate sample buffer and NuPAGE reducing agent (Invitrogen), resolved by 4% to 12% Bis/Tris NuPAGE gel Immunohistochemistry Study (Invitrogen), transferred onto Hybond-C extramembrane Immunohistochemical staining was done on 4-μm-thickar- (Amersham Biosciences), and immunoblotted with the indicat- chival formalin-fixed paraffin-embedded tissue sections. Sec- ed antibodies as described previously (23). tions were deparaffinized twice in xylene for 10 min and Genotoxin-treated cells were harvested, washed with PBS, twice in ethanol for 2 min and then placed in 100 mmol/L Tris, fixed with 70% ethanol for 2 h, or stored at −20°C. The 50 mmol/L EDTA (pH 6.0) buffer and heated at 92°C for fixed samples were resuspended in PBS with 10 μg/mL DN- 15 min. Samples were washed and the endogenous peroxidase ase-free RNase A (Sigma) and 10 μg/mL propidium iodide activity was blocked by 30% H2O2 for 5 min. The CD31 mono- (Sigma) at 4°C. Cell cycle profiling was analyzed by BD clonal antibody (BD Pharmingen) was incubated for 2 h at FACSCalibur flow cytometer (Becton Dickinson) and pro- room temperature, washed with PBS, and then incubated with cessed with ModFit software, version 2.0 (Verity Software the Dako LSAB2+ HRP System (DakoCytomation Denmark House). Polyploidy populations were quantified with WinM- A/S) and counterstained with hematoxylin. Capillary density DI software, version 2.8. indicated by CD31 staining was detected by a Nikon Opti- phot-2 upright microscope at 20× objective and analyzed using Cell Adhesion Assay Northern Eclipse software (Empix Imaging). Ninety-six–well plates were coated with laminin (10 μg/mL) and incubated at room temperature for 1 h, subsequently washed with PBS, and blocked with 1% bovine serum albumin Disclosure of Potential Conflicts of Interest No potential conflicts of interest were disclosed. at 37°C for 1 h. MCF-10A cells (2 × 106/mL) were suspended in DMEM/F12 serum-free medium. Cell suspensions (100 μL) were added into each well and incubated in 5% CO2 at 37°C for Acknowledgments 1 h. The nonadherent cells were removed by PBS washing. We thankDr. Ronald B. Gartenhaus for much help and Dr. Ning-Hsing Yeh for the gift of NuMA monoclonal antibody. Adherent cells were fixed with methanol for 15 min and then stained with 0.2% crystal violet in 2% ethanol for another 30 min. Adherent cells were rinsed, lysed with 0.2% Triton References for 30 min, and measured by spectrophotometer at 595-nm 1. BozcukH, Uslu G, Pe°tereli E, et al. Predictors of distant metastasis at presentation in breast cancer: a study also evaluating associations among common absorbance. biological indicators. Breast Cancer Res Treat 2001;68:239–48. 2. Borel F, Lohez OD, Lacroix FB, Margolis RL. Multiple centrosomes arise from Cell Migration Assay tetraploidy checkpoint failure and mitotic centrosome clusters in p53 and RB pocket Polyethylene terephthalate–coated membrane (8-μmpore protein-compromised cells. Proc Natl Acad Sci U S A 2002;99:9819–24. size, HTS FluoroBlokInsert, Falcon) was placed onto 24-well 3. Salisbury JL, D'Assoro AB, Lingle WL. Centrosome amplification and the

origin of chromosomal instability in breast cancer. J Mammary Gland Biol Neo- 20. Knauf JA, Ouyang B, Knudsen ES, Fukasawa K, Babcock G, Fagin JA. On- plasia 2004;9:275–83. cogenic RAS induces accelerated transition through G2/M and promotes defects in the G2 DNA damage and mitotic spindle checkpoints. J Biol Chem 2006;281: 4. Fukasawa K. Centrosome amplification, chromosome instability and cancer – development. Cancer Lett 2005;230:6–19. 3800 9. 5. van de Wetering CI, Horne MC, Knudson CM. Chromosomal instability and 21. Shi B, Hsu HL, Evens AM, Gordon LI, Gartenhaus RB. Expression of the supernumerary centrosomes represent precursor defects in a mouse model of candidate MCT-1 oncogene in B- and T-cell lymphoid malignancies. Blood 2003; – T-cell lymphoma. Cancer Res 2007;67:8081–88. 102:297 302. 6. Brunet S, Vernos I. Chromosome motors on the move. From motion to spindle 22. Hsu HL, Shi B, Gartenhaus RB. The MCT-1 oncogene product impairs cell checkpoint activity. EMBO Rep 2001;2:669–73. cycle checkpoint control and transforms human mammary epithelial cells. Onco- gene 2005;24:4956–64. 7. Li JJ, Li SA. Mitotic kinases: the key to duplication, segregation, and cytokinesis errors, chromosomal instability, and oncogenesis. Pharmacol Ther 2006; 23. Hsu HL, Choy CO, Kasiappan R, et al. MCT-1 oncogene downregulates p53 111:974–84. and destabilizes genome structure in the response to DNA double-strand damage. DNA Repair 2007;6:1319–32. 8. Malumbres M, Barbacid M. Cell cycle kinases in cancer. Curr Opin Genet Dev 2007;17:60–5. 24. ProsniakM, Dierov J, OkamiK, et al. A novel candidate oncogene, MCT-1, is involved in cell cycle progression. Cancer Res 1998;58:4233–7. 9. Jallepalli PV, Lengauer C. Chromosome segregation and cancer: cutting through the mystery. Nat Rev Cancer 2001;1:109–17. 25. Nandi S, Reinert LS, Hachem A, et al. Phosphorylation of MCT-1 by p44/42 10. Draviam VM, Xie S, Sorger PK. Chromosome segregation and genomic sta- MAPK is required for its stabilization in response to DNA damage. Oncogene 2007;26:2283–9. bility. Curr Opin Genet Dev 2004;14:120–5. 11. Weaver BA, Cleveland DW. Decoding the links between mitosis, cancer, and 26. Fleischer TC, Weaver CM, McAfee KJ, Jennings JL, LinkAJ. Systematic identification and functional screens of uncharacterized proteins associated with chemotherapy: the mitotic checkpoint, adaptation, and cell death. Cancer Cell – 2005;8:7–12. eukaryotic ribosomal complexes. Genes Dev 2006;20:1294 307. 12. Fukasawa K, Choi T, Kuriyama R, Rulong S, Vande Woude GF. Abnormal 27. Reinert LS, Shi B, Nandi S, et al. MCT-1 protein interacts with the cap com- centrosome amplification in the absence of p53. Science 1996;271:1744–7. plex and modulates messenger RNA translational profiles. Cancer Res 2006;66: 8994–9001. 13. Sphyris N, Harrison DJ. p53 deficiency exacerbates pleiotropic mitotic de- 28. Levenson AS, Thurn KE, Simons LA, et al. MCT-1 oncogene contributes to fects, changes in nuclearity and polyploidy in transdifferentiating pancreatic aci- in vivo nar cells. Oncogene 2005;24:2184–94. increased tumorigenicity of MCF7 cells by promotion of angiogenesis and inhibition of apoptosis. Cancer Res 2005;65:10651–6. 14. Yun C, Cho H, Kim SJ, et al. Mitotic aberration coupled with centrosome 29. Wakisaka N, Pagano JS. Epstein-Barr virus induces invasion and metastasis amplification is induced by hepatitis B virus X oncoprotein via the Ras-mitogen- – activated protein/extracellular signal-regulated kinase-mitogen-activated protein factors. Anticancer Res 2003;3:2133 8. pathway. Mol Cancer Res 2004;2:159–69. 30. Giancotti FG. Targeting integrin β4 for cancer and anti-angiogenic therapy. – 15. Fukasawa K, Vande Woude GF. Synergy between the Mos/mitogen-activated Trends Pharmacol Sci 2007;28:506 11. protein kinase pathway and loss of p53 function in transformation and chromo- 31. Nikolopoulos SN, Blaikie P, Yoshioka T, Guo W, Giancotti FG. Integrin β4 some instability. Mol Cell Biol 1997;17:506–18. signaling promotes tumor angiogenesis. Cancer Cell 2004;6:471–83. 16. Saavedra HI, Fukasawa K, Conn CW, Stambrook PJ. MAPK mediates RAS- 32. Kanies CL, Smith JJ, Kis C, et al. Oncogenic Ras and transforming growth induced chromosome instability. J Biol Chem 1999;274:38083–90. factor-β synergistically regulate AU-rich element-containing mRNAs during ep- – 17. Saavedra HI, Knauf JA, Shirokawa JM, et al. The RAS oncogene induces ithelial to mesenchymal transition. Mol Cancer Res 2008;6:1124 36. genomic instability in thyroid PCCL3 cells via the MAPK pathway. Oncogene 33. Halaschek-Wiener J, Wacheck V, Kloog Y, Jansen B. Ras inhibition leads to 2000;19:3948–54. transcriptional activation of p53 and down-regulation of Mdm2: two mechanisms that cooperatively increase p53 function in colon cancer cells. Cell Signal 2004; 18. Shin I, Kim S, Song H, Kim HR, Moon A. H-Ras-specific activation of Rac- – MKK3/6-p38 pathway: its critical role in invasion and migration of breast epithe- 16:1319 27. lial cells. J Biol Chem 2005;280:14675–83. 34. Mazan-Mamczarz K, Hagner PR, Corl S, et al. Post-transcriptional gene regulation by HuR promotes a more tumorigenic phenotype. Oncogene 2008;27: 19. Lee KW, Kim MS, Kang NJ, et al. H-Ras selectively up-regulates – MMP-9 and COX-2 through activation of ERK1/2 and NF-κB: an implica- 6151 63. tion for invasive phenotype in rat liver epithelial cells. Int J Cancer 2006; 35. Jänicke RU, Sohn D, Schulze-Osthoff K. The dark side of a tumor suppres- 119:1767–75. sor: anti-apoptotic p53. Cell Death Differ 2008;15:959–976, Review.

Mol Cancer Res 2009;7(4). April 2009 Downloaded from mcr.aacrjournals.org on September 24, 2021. © 2009 American Association for Cancer Research. Loss of p53 and MCT-1 Overexpression Synergistically Promote Chromosome Instability and Tumorigenicity

Ravi Kasiappan, Hung-Ju Shih, Kang-Lin Chu, et al.

Mol Cancer Res 2009;7:536-548.

Updated version Access the most recent version of this article at: http://mcr.aacrjournals.org/content/7/4/536

Supplementary Access the most recent supplemental material at: Material http://mcr.aacrjournals.org/content/suppl/2009/04/30/7.4.536.DC1

Cited articles This article cites 35 articles, 14 of which you can access for free at: http://mcr.aacrjournals.org/content/7/4/536.full#ref-list-1

Citing articles This article has been cited by 3 HighWire-hosted articles. Access the articles at: http://mcr.aacrjournals.org/content/7/4/536.full#related-urls

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

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mcr.aacrjournals.org/content/7/4/536. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.