OverexpressionofcyclinB1inhumanesophagealsquamouscellcarci ... · Adhesion assay Tumor cell–endothelial cell adhesion assay was performed as described (31,32). In total, 96-well

Carcinogenesis vol.29 no.2 pp.307–315, 2008doi:10.1093/carcin/bgm269Advance Access publication November 28, 2007

Overexpression of cyclin B1 in human esophageal squamous cell carcinoma cells inducestumor cell invasive growth and metastasis

Yongmei Song, Chunling Zhao, Lijia Dong, Ming Fu, LiyanXue, Zhen Huang, Tong Tong, Zhuan Zhou, Amei Chen,Zhihua Yang, Ning Lu and Qimin Zhan�

State Key Laboratory of Molecular Oncology, Cancer Institute and Hospital,Chinese Academy of Medical Sciences and Peking Union Medical College,Beijing 100021, China

�To whom correspondence should be addressed. Tel: þ86 10 67762694;Fax: þ86 10 67715058;Email: [email protected]

Cyclin B1, a key component in the control of cell cycle progressionfromG2 to M phase, has been implicated in tumorigenesis and thedevelopment of malignancy. However, the underlying mechanismby which cyclin B1 acts as an important oncogenic molecule re-mains largely unknown. Here we show that ectopic expression ofcyclin B1 promotes cell proliferation, enhances cell motility andmigration and results in increased ability of cells extravasatingthrough the capillary endothelium. Interestingly, isogenic esoph-ageal squamous cell carcinoma (ESCC) cells overexpressing cy-clin B1 reveal strong invasive growth and high potential ofmetastasis to lung in xenograft mice. Suppression of cyclin B1expression via small interfering RNA approach in high-metastaticesophagus carcinoma cells specifically inhibits their ability to me-tastasize from the primary ESCC to lung. Notably, altered expres-sion of epithelial markers and mesenchymal markers wereobserved in the cells overexpressing cyclin B1, suggesting thatcyclin B1 contributes to metastasis probably by promoting anepithelial–mesenchymal transition. These results establish amech-anistic link between cyclin B1 and ESCC metastasis and providenovel insight into understanding of cyclin B1 in the developmentof ESCC malignancy.

Introduction

Tumor metastasis is the most common cause of death in cancer pa-tients. The metastatic process is thought to consist of a number ofdistinct steps, including tumor cell local invasion, intravasation, sur-vival in circulation, arrest in capillaries, extravasation and finally out-growth to produce macrometastasis at distant organs (1–3). Therefore,the metastatic cascade is an ordered sequence of events required formetastasis to occur. It has been demonstrated that a number of mo-lecular pathways and cellular mechanisms that underlie the multistageprocess of metastasis formation. Multiple signal transduction path-ways, changes in the adhesive and migratory capabilities of tumorcells and the tumor microenvironment have critical roles in malignanttumor progression (4). For example, matrix metalloproteinase (MMP)3 expression results in transcriptional upregulation of the small gua-nosine triphosphatase Rac1b and upregulated levels of reactiveoxygen species, which, in turn, induces expression of Snail, loss ofE-cadherin expression, genomic instability and tumor progression (5).MMPs can also promote tumor cell invasion and metastatic dissem-ination by activating the protease-activated receptor 1, a G-protein-coupled receptor implicated in metastasis of various cancer types (6).Recent gene expression profiling experiments with breast cancer celllines metastasizing to specific target organs have revealed a list ofinteresting genes and factors. Genes that direct breast cancer cells to

the lung include secreted protein, acidic, cysteine rich, inhibitor ofDNA binding 1, MMP1, MMP2, vascular cell adhesion molecule 1,interleukin 13 receptor-a, cyclooxygenase 2 and CXCL1 (7). RhoCwas reported to induce lung metastasis of melanoma cells, and in-terleukin 11 and osteopontin expressions were found to be upregu-lated in breast cancer cells metastasizing to bone (8,9). Anotherexample of integrin-mediated prometastatic signaling was providedby the discovery of periostin, a protein that is highly expressed inmetastatic colorectal cancer (10). Additionally, several lines of evi-dence have demonstrated that epithelial–mesenchymal transition(EMT) is tightly linked to tumor metastasis. Yang et al. (11) reportthat the transcription factor Twist, whose expression induces EMT inepithelial cells, is an important regulator in the process of metastasis.Suppression of Twist expression in tumor cells greatly inhibits themetastatic potential. This study provides direct evidence that EMTmay be an essential process during metastasis.

Human esophageal squamous cell carcinoma (ESCC) is one of themost frequent malignancies worldwide and occurs at a very highfrequency in China, South Africa, France and Italy (12). ESCC isa highly aggressive neoplasm. On a global basis, cancer of the esoph-agus is the sixth leading cause of cancer death worldwide and fourthleading death cause in China. With advances in surgical techniquesand treatment, the prognosis of esophageal cancer has slowly im-proved over the past three decades. However, the 5-year overall sur-vival rate (14%) remains poor, even in comparison with the dismalsurvival rates (4%) from the 1970s. The underlying reasons for thisdisappointingly low survival rate are multifold. Nonetheless, distantmetastasis remains the primary cause of death. With the addition ofnovel targeted therapies, the goal is to improve the response rate andreduce distant metastasis without significant additive side effects (13).

Cyclin B1 plays an important role as a mitotic cyclin in G2/M phasetransition during the cell cycle. In the normal cell cycle, it starts toexpress at the late S phase, increase in G2 phases and peak at mitosis.Cyclin B1 associates with p34cdc2, in common with other cyclins, thecyclin B1 possess a conserved 100 amino acid region called the ‘cy-clin box’ (14), which binds to the cyclin-dependent kinase. Theirassociated kinase activities appear when cells enter mitosis and dis-appear as the cyclins are destroyed in anaphase. At the end of themetaphase, a specialized device, the anaphase- promoting complexmust destroy cyclin B1 to allow mitosis to proceed. The anaphase-promoting complex, a polyprotein complex activated by Cdc20 re-cruits cyclin B1, causes its ubiquitination and, thus, targets it fordegradation by the 26S proteasome (15). Therefore, the expressionof cyclin B1 is strictly dependent on cell cycle progression. It has beenrespected that deregulated cyclin B1 may probably lead to uncon-trolled cell cycle progression and cause cell malignant transformation.Indeed, overexpression of cyclin B1 has been observed in varioushuman tumors, including ESCC, lung cancer, squamous cell carci-noma of the head and neck, breast cancer, colon adenocarcinoma,renal cell carcinoma, hepatocellular carcinoma, pancreatic cancer,laryngeal squamous cell carcinoma, prostate cancer, cervical carci-noma, colorectal cancer, astrocytomas and oral carcinoma (16–23).Interestingly, it has been suggested that deregulated cyclin B1 expres-sion might be used as an indicator of the malignant potential of thetumors (18–22). However, the mechanism between cyclin B1 andtumor progression remains to be greatly elucidated.

In this report, using cyclin B1-expressing human ESCC cells, wedemonstrated that overexpression of cyclin B1 promotes cell invasivegrowth and extravasation through the capillary endothelium. In xeno-graft mice, overexpression of cyclin B1 is probably able to enhance lungmetastasis of ESCC, and suppression of endogenous cyclin B1 inhibitsmetastatic potential of ESCC to lung. Furthermore, cyclin B1-inducedESCC metastasis appears to be related to alteration of EMT molecules.

Abbreviations: ECEC, esophageal carcinoma endothelium cell; EMT,epithelial–mesenchymal transition; ESCC, esophageal squamous cell carci-noma; HLEC, human lung endothelium cell; MMP, matrix metalloproteinase;NF-jB, nuclear factor kappa B; PBS, phosphate-buffered saline; TEM, trans-endothelial migration.

� The Author 2007. Published by Oxford University Press. All rights reserved. For Permissions, please email: [email protected] 307

at Pennsylvania State University on February 23, 2013

http://carcin.oxfordjournals.org/D

ownloaded from

http://carcin.oxfordjournals.org/

Methods

Cell culture

Human ESCC cell lines were grown in RPMI 1640 medium supplemented with10% fetal bovine serum at 37�C under 5% CO2 and saturated moisture.EC9706 cell line was generously provided by Dr Mingrong Wang in CancerInstitute (Hospital), Peking Union Medical College, Chinese Academy ofMedical Sciences. Endothelial cell human lung endothelium cell (HLEC)and esophageal carcinoma endothelium cell (ECEC) were isolated and culturedas described (24–26).

small interfering RNA (siRNA) studies

The siRNA sequence that we used are listed as follows: cyclin B1, ATGAGAGC-CATCCTAATTG (27), and control siRNA, UUCUCCGAACGUGUCACGU.Cells were transfected with cyclin B1 siRNA or control siRNA using Lipofect-amine 2000 (Invitrogen Corporation, Carlsbad, CA) according to the manufac-turer’s instruction. The cells were cultured in the presence of 400 lg/ml of G418(Geneticin sulfate; GIBCO) for 2 weeks. For each transfection, all the colonieswere trypsinized and collected to produce stable cell pools.

Stable transfections and establishment of stable cell lines

Transfection was performed in 80% confluent cells using 8 ll of Lipofectamine2000 (Invitrogen Corporation) and 5 lg of pcDNA3.1 cyclin B1 or 5 lg ofpcDNA3.1 empty vector. Stable transfectants were obtained by 400 lg/ml ofG418 selection for �14 days.

Cell growth curve

In total, 2 � 104 cells were plated in 35 mm tissue culture dishes (NUNC,Thermo Fisher Scientific, Waltham, MA) in RPMI 1640 with 10% fetal bovineserum. The cells were trypsinized and manually counted on days 1, 2, 3, 5and 7. For each plate, cell count was repeated three times to draw the cellgrowth curve.

Colony formation assay

In 100 mm tissue culture dishes, 1 � 103 cells were grown. After 14 days, the cellswere washed with phosphate-buffered saline (PBS) and fixed with methanol and0.1% crystal violet. The colonies were manually counted and then photographed.

Invasion assay

All procedures were performed as described (28). Invasion of cells was assayedin Transwell cell culture chambers with 6.5 mm diameter polycarbonate mem-brane filters containing 8 lm pore size (Neuro Probe, Gaithersburg, MD). Themembrane was coated with 20 ml of a 2.5 mg/ml solution of matrigel (FalconBD, San Jose, CA). In total, 1 � 104 cells in 50 ll of culture medium wereadded to the upper chamber of the device, and the lower chamber was filledwith 30 ll of medium. After 24 h of incubation at 37�C, the non-invasion cellswere removed from the upper surface of the membrane with a cotton swab. Thefilters were then fixed in methanol for 10 min, stained with Giemsa solution for1 h and counted. Ten random microscopic fields (�400) were counted per welland the mean was determined.

Transmigration assay (transendothelial migration assay)

Transendothelial migration (TEM) assay was performed as described (29,30). Inbrief, endothelial cells were seeded into 24-well transwell 8 lm pore polycar-bonate membrane (costar) and further cultured for 5 days to grow into non-permeable endothelial monolayer. Then, 4 � 104 tumor cells labeled withpKH26 (Sigma, St Louis, MO) were added to the apical chamber and incubatedfor more 24 h. After removing non-migrating cells, the transmigrating cellswere counted from 10 random fields of �200 magnification under a fluorescentmicroscope.

Adhesion assay

Tumor cell–endothelial cell adhesion assay was performed as described(31,32). In total, 96-well plates were coated with 1% gelatin (Sigma) inD-Hanks and then aspirated after 30 min incubated at 37�C. In total, 9 �103 ECEC and 1.2 � 104 HLEC were seeded per well and cultured in endo-thelial growth medium until grown to confluent monolayer, then replaced thegrowth medium to quiescent medium (M199 medium with 0.1% fetal bovineserum, free endothelial cell growth supplement) and, subsequently, cultured for3 more days to mimic the quiescent endothelium in vivo. Tumor cells werespecifically labeled with Calcein AM (Molecular Probes) according to themanufacturer’s instruction and then incubated with endothelial monolayerfor 1 h at 37�C without shaking. Following this incubation, endothelial mono-layers were washed four times with RPMI 1640 medium with 0.5% bovineserum albumin to remove non-adherent cells. The adherent tumor cells werephotoed by fluorescent microscopy under �100 magnification and counted byIPP5.1plus soft. For studying the tumor cell-stimulated tumor–endothelial ad-

hesion, when the endothelial cell monolayer grown to subconfluence, replacedthe growth medium to tumor cell conditional medium, and subsequentlyculture for more 2 days before adhesion assay. For preparing the tumor cellconditional medium, when the tumor cell were grown to 80%, replaced thegrown medium to conditional medium (RPMI 1640 without fetal bovine me-dium), and subsequently culture for more 2 days. Then the conditional mediumwas centrifuged to deplete cell debris and stored at �20�C until use.

Mice and injection

BALB/c nude mice, 4 weeks old, were provided by the Cancer Institute for thein vivo tumorigenicity study. The experiment was performed along established,institutional animal welfare guidelines concordant with National Institutes ofHealth species criteria. Mice were injected subcutaneously with 1 � 106 cellsin 0.1 ml into the right upper back and raised in the following 90 days. Themice were then monitored for tumor volume, overall health and total bodyweight. The size of the tumor was determined by caliper measurement of thesubcutaneous tumor mass. Tumor volume was calculated according to theformula 4/3pr1

2r2 (r1 , r2). Each experimental group contained 10 mice. Atthe end of 3 months, all mice were killed, and the tumor volume and weightwere measured. All the tumors were fixed in 4% polyformaldehyde and cut intoconsecutive 6 lm sections. For observation of metastasis, all the lungs, livers,spleens, kidneys, stomachs and brains were excised; fixed in polyformaldehydeand cut into consecutive 6 lm sections. All these sections above were stainedwith hematoxylin and eosin and observed under a microscope. The presence oflung metastasis was determined and confirmed by the pathologists in a cancerhospital (Chinese Academy of Medical Sciences). Three independent experi-ments were performed and gave similar results.

Confocal imaging

Cells were grown on coverslips in RPMI 1640 complete medium for 24 h. Thecoverslips were rinsed thrice with ice-cold PBS and, simultaneously, fixed andpermeabilized by immersion in cold methanol (�20�C) for 15 min. Coverslipswere rinsed with PBS and inverted onto a 100 ll of PBS with the correspondingantibodies (1:250). After overnight incubation with antibody at 4�C and sub-sequent wash with PBS, coverslips were inverted again onto 100 ll of PBS con-taining the corresponding Tetramethyl Rhodamine Isothiocyanate-conjugatedsecondary antibody (1:100) and incubated 30 min at room temperature. Cover-slips were washed in PBS and counterstained with 4#,6#-diamidino-2-phenyl-indole (0.1 lg/ml). Slides were examined with an ultraspectral confocalmicroscopy system. Series of images were processed and analyzed with theaccompanying software package.

Results

Cyclin B1 enhances cell proliferation and invasion in vitro

A number of investigations indicate that deregulation of cyclin B1 isclosely associated with ESCC (19,33). To understand the biologicalrole of cyclin B1 in ESCC progression, we established cyclin B1-expressing cell lines. ESCC KYSE150 cells, which had relativelylow levels of endogenous cyclin B1 expression (Figure 1A), werestably transfected with a cyclin B1 expression construct or with vectoralone as the control (Figure 1B). Two cyclin B1-overexpressing clonesand one vector control clone were used for further analyses(Figure 1C). We designated these two clones high-CycB1-1 andhigh-CycB1-2, whereas the vector-transfected clone was termedpcDNA3.1. Compared with the relatively low levels of endogenouscyclin B1 in parental KYSE150, high-CycB1-1 and high-CycB1-2expressed at least 3- to 5-fold more cyclin B1 (Figure 1C). We firstexamined the effect of cyclin B1 on cell proliferation and found thatgrowth of high-CycB1-1and high-CycB1-2 cells significantly in-creased compared with the pcDNA3.1 cells or parental KYSE150 cells(Figure 1D). Next, cell matrigel assay was performed to evaluate theinvasion of these cell lines since cell migrating ability is closely associ-ated with the potential of invasive growth. As shown in Figure 1E, high-CycB1-1 and high-CycB1-2 cells exhibited stronger invasion, whereasboth pcDNA3.1 and KYSE150 displayed weak invasive ability. Takentogether, these results indicated that cyclin B1-overexpressing cellsstrongly enhanced cell proliferation and invasion in vitro, suggestingthat overexpression of cyclin B1 enhance migration of ESCC cells andmight contribute to the development of tumor invasive growth.

Cyclin B1 promotes tumor growth in vivo

To gain further support for cyclin B1 contributing to ESCC develop-ment, an in vivo mouse model was employed. Squamous cell

Y.Song et al.

308



ownloaded from


carcinoma had been found in the xenograft mice (Figure 2B). High-CycB1-1, high-CycB1-2, pcDNA3.1 and KYSE150 were implantedsubcutaneously into the right upper back of BALB/c mice. They

formed ESCC carcinomas within 2 weeks. Our data showed that thetumors from cyclin B1-transfected cells in nude mice grew morerapidly than that from KYSE150 or empty vector cells (Figure 2A),

Fig. 1. Cyclin B1 induces proliferation and invasion of KYSE150 in vitro. (A) Twelve cell lines of ESCC were grown in RPMI 1640 10% fetal bovine serum. Intotal, 100 lg of whole-cell protein was used for western blot analysis with anti-cyclin B1 antibody as described. (B) KYSE150 cells were transfected withpcDNA3.1-cyclin B1-expressing vector; clones expressing a transfected cyclin B1 protein were detected by western blot analysis. (C) High-CycB1-1, high-CycB1-2, pcDNA3.1 and KYSE150 were detected by western blot analysis. b-Actin was used as an equal loading control for western blot. (D) Growth curvesreveal that high-CycB1 cells proliferate more rapidly than pcDNA3.1 and KYSE150 cells. (E) Cyclin B1 expression significantly induces invasion of KYSE150cells through a fibronectin-coated transwell fiber. Invasion assay was quantified by counting the invasion cells in 10 random high-powered fields per filter. Allexperiments were performed at least three times with consistent and repeatable results.

Fig. 2. Cyclin B1 promotes the ability of KYSE150 cells to invade the adjacent tissues and metastasize from the primary ESCC to lung in nude mice. (A) Thevolume of tumor was calculated according to the formula 4/3pr1

2r2 (r1 , r2). Three independent experiments were performed. (B) Histological analysis ofsquamous cell carcinoma. (C) Representative hematoxylin and eosin staining sections of invasional bone formed by high-CycB1 cells (invasional bone).Representative hematoxylin and eosin staining sections of invasional striated muscle formed by high-CycB1 cells (invasional striated muscle). Representativehematoxylin and eosin staining sections of metastatic nodules in the lung formed by high-CycB1 cells.

Overexpression of cyclin B1 in human ESCC cells

309



ownloaded from


which was consistent with the cell growth curve above. At the end of 3months, the weights of tumors from cyclin B1-transfected cells werefour to five times more than that of tumors from KYSE150 or emptyvector cells.

Cyclin B1 promotes tumor invasion and metastasis in nude mice

Mice were killed at the end of 3 months for examination of tumorinvasion and metastasis. Interestingly, these four cell lines differeddramatically in their metastatic potentials. Apparently, parentalKYSE150 and pcDNA3.1 cell lines only formed in situ primary tu-mors, but no metastatic tumors were detected in any distant tissues,such as lymph nodes, lung, liver, brain, kidney and spleen. In contrast,high-CycB1-1 and high-CycB1-2 lines were probably able to com-plete all steps of metastasis and formed visible metastatic nodules inlung efficiently (Figure 2C and Table I). Interestingly, the cyclin B1-associated metastasis appeared to specifically occur in lung, but un-likely in other tissues or organs. Consistent with these findings, thesetwo lines were observed to invade into adjacent tissues such as bone orstriated muscle (Figure 2C). Therefore, ESCC cell expressing highlevels of cyclin B1 is able to promote tumor invasive growth in vivoand probably results in lung metastasis.

Suppression of cyclin B1 expression reduces cell proliferation, colonyformation and invasion

To further determine whether cyclin B1 plays an essential role inESCC metastasis, we tested whether inhibition of cyclin B1 expres-sion in the high-metastatic EC9706 cells would affect their metastaticability. To do so, we stably transfected pGCsi-U6/Neo/GFP/shCyclinB1 (a cyclin B1 siRNA expression vector) into EC9706 cells, anESCC cell line with high malignancy and mainly generates lungmetastasis (28), and selected two clones (CycB1-siRNA1 andCycB1-siRNA2) for further experiments. Evidently, the protein levelsof endogenous cyclin B1 in both CycB1-siRNA1 and CycB1-siRNA2were reduced by .70%, but cells transfected with control siRNAdid not show any significant reduction of cyclin B1 expression(Figure 3A).

Both CycB1-siRNA cells grew slower and revealed decreased col-ony formation than that seen in control siRNA cells (Figure 3B and D).For the colony formation assay, the rate of colony growth was moreobviously decreased in the EC9706–cyclin B1 siRNA cells. The col-onies from EC9706–cyclin B1 siRNA cells were much smaller thanthose from the EC9706–control siRNA cells. Additionally, the twoCycB1-siRNA cell lines exhibited weak invasion ability comparedwith EC9706 and control siRNA cells in transwell assays (Figure3C). These results suggested that suppression of cyclin B1 expressionreduces cell proliferation, colony formation and migration in vitro.

Cyclin B1 greatly contributes to lung metastasis of ESCC

Next, we tested whether inhibition of cyclin B1 expression in EC9706cells would affect their metastatic ability in vivo. EC9706–cyclin B1

siRNA cells and EC9706–control siRNA cells were implanted in theright upper back of BALB/c mice, and the animals were killed at theend of 3 months. All the lungs, livers, spleens, kidneys and brainswere cut into consecutive 6 lm sections. The metastasis was carefullyobserved on serial microscopic sections of whole specimens in allmice. Lung metastases were seen in 3 of 10 mice from EC9706 cellswithout the help of a microscope. By a dissection microscope, paren-tal or empty vector EC9706 cells metastasized to the lungs in half ofnude mice, resulting in a total number of 28 or 29 lung tumor colonies,respectively. Whereas the cells expressing control siRNA formedlarge numbers of macroscopically visible metastases in their lungs,those that expressed CycB1-siRNA exhibited very few measurablemetastases (Table I). Furthermore, the primary tumor volumes in situof CycB1-siRNA cells were all markedly smaller than those from thecontrol siRNA and parental EC9706 (Figure 3E). The results sug-gested that suppression of endogenous cyclin B1 by RNAi in meta-static cancer cells significantly decreases the efficiency of lungmetastasis from the primary tumor, suggesting that cyclin B1 is im-portant for ESCC metastasis from the primary tumor to lung.

Overexpression of cyclin B1 in ESCC cells promotes TEM

Cancer metastasis is a complex multiple process, extravasation is oneof the least known steps in cancer metastasis and it involves dynamicinteractions between cancer cells and the endothelium (34). We nextinvestigated the effect of cyclin B1 on tumor cell TEM. Clearly, thehigh-CycB1-1 and high-CycB1-2 cell lines showed a stronger abilityin transmigrating through the ECEC than pcDNA3.1 and KYSE150cells (Figure 4A and B). Especially, we performed the same experi-ments in normal HLEC as it was done in ECEC and found that high-CycB1-1 and high-CycB1-2 cell lines transmigrated strongly thanpcDNA3.1 and KYSE150 cells (Figure 4C and D). Additionally, wealso examined the adhesion of those isogenic cell lines to ECEC andHLEC, but did not find the significant difference between cyclin B1-expressing ESCC cells and control or parental ESCC cells (Figure 4Eand F). These results indicate that overexpression of cyclin B1 pro-motes ESCC intravasation in ECEC and extravasation in HLEC, butdoes not alter ESCC adhesion to those cells (ECEC and HLEC).

Cyclin B1-enhanced metastasis is associated with induction of EMT

Recently, it has been reported that the breakdown of epithelial cellhomeostasis leading to aggressive cancer progression has been corre-lated with the loss of epithelial characteristics and the acquisition ofa migratory phenotype (35). This phenomenon, referred to as EMT, isconsidered a crucial event in late-stage tumorigenesis (36). We there-fore speculated whether the contribution of cyclin B1 to tumor me-tastasis might involve the induction of an EMT. To determine whetherthe molecular alterations of an EMT occurred in the cells expressingcyclin B1, we examined the localization of adhesion junction proteinssuch as E-cadherin and b-catenin. Immunofluorescent assays showedthat the two proteins substantially reduced in the membrane of high-

Table I. The invasion and metastasis in nude mice comparison between cell lines of cyclin B1 different expression

Cell lines

KYSE150 pcDNA3.1 High-CycB1-1 High-CycB1-2 EC9706 Control siRNA CycB1 siRNA1 CycB1 siRNA2

Bone 0 0 3 6 0 0 0 0Striated muscle 2 3 9 11 0 0 0 0Lung 1 0 17 19 16 17 1 2Liver 0 0 0 0 0 0 0 0Stomach 0 0 0 0 0 0 0 0Brain 0 0 0 0 0 0 0 0Spleen 0 0 0 0 0 0 0 0Kidney 0 0 0 0 0 0 0 0Lymph 2 1 9 10 12 13 2 4Total case 30 30 30 30 30 30 30 30

Y.Song et al.

310



ownloaded from


cyclin B1 cells, compared with the strong staining of E-cadherin andb-catenin in the membrane of the control cells (Figure 5B). Similarresults were also observed in immunoblotting analysis (Figure 5A).This finding underlines the observation that cyclin B1 induces a cad-herin switch. In contrast, the expression of fibroblast markers, includ-ing fibronectin and N-cadherin, whose expression has been shown tocorrelate positively with the EMT (37), was strongly induced in thehigh-CycB1 cells (Figure 5A and B). Hence, both the morphologicaland molecular changes in the cyclin B1-expressing ESCC cells dem-onstrated that these cells had undergone an EMT. We also examinedthe expression of nuclear factor kappa B (NF-jB) due to its greatcontribution to EMT and found that NF-jB expression level wasupregulated in high-CycB1 cell lines, suggesting that deregulationof cyclin B1 probably activates the NF-jB signaling pathway (Figure5A and B). Furthermore, we detected two other EMT markers, Snailand Twist. Twist showed parallel trend with the transition from epi-thelium to mesenchyma. (Figure 5A). Unfortunately, the Snail is ex-pressed at extremely low level and almost undetectable. To furtherdetermine whether cyclin B1 plays a crucial role in EMT, we exam-ined whether inhibition of cyclin B1 expression in ESCC cell lineswould affect their transition. To do so, we chose three cell lines with

relatively high levels of endogenous cyclin B1 expression and trans-fected these cell lines with pGCsi-U6/Neo/GFP/shCyclin B1 (a cyclinB1 siRNA expression vector) and examined two EMT markers, E-cadherin and Twist. We observed that the expression of E-cadherinrevert along with reduction of cyclin B1 expression. In contrast, thelevel of Twist expression decreases following suppressed cyclin B1expression by cyclin B1 siRNA (Figure 5A).

Discussion

Altered regulation of the cell cycle is one of the hallmarks of humancancer (38). Uncontrolled cell division is an indispensable event intumor progression, and numerous molecules involved in this processhave been the focus of intense investigation in tumor biology. Anumber of investigations have shown that cyclins, molecules thatorchestrate normal cell cycle progression, are abnormally overex-pressed in various human cancers (39). Cyclin B1, an essential cellcycle regulator, is remarkably overexpressed in human ESCC tissuescompared with their normal adjacent tissues.

We have conducted a series of experiments to explore the molecularmechanisms involved in cyclin B1-related malignant phenotypes of

Fig. 3. Suppression of cyclin B1 expression inhibits proliferation and invasion of EC9706 cells in vitro. (A) EC9706 cells were transfected with either cyclin B1siRNA or control siRNA. Two CycB1-siRNA clones designated as CycB1-siRNA1 and CycB1-siRNA2, one control siRNA and EC9706 parental cells weredetected by western blot analysis and reverse transcriptase–polymerase chain reaction. b-Actin was used as an equal loading control for western blot andglyceraldehyde-3-phosphate dehydrogenase was provided to document equivalent loading for reverse transcriptase–polymerase chain reaction. (B) Growth curvesreveal that CycB1-siRNA1 and CycB1-siRNA2 cells proliferate rather slower than control siRNA and EC9706 cells. (C) Analysis of cell invasion ability. In total,1 � 104 cells were used for transwell migration assay. Migration was quantified by counting the migrated cells in 10 random high-powered fields per filter. (D)Plating efficiency assay reveals changes in colony formation ability. All experiments were performed at least three times with consistent and repeatable results. (E)Tumor growth curve from EC9706 cells. The volume of tumor was calculated according to the formula 4/3pr1

2r2 (r1 , r2). Three independent experiments wereperformed.


311



ownloaded from


ESCC. Through employment of isogenic cells expressing cyclin B1,we have demonstrated that cyclin B1 protein substantially promotesthe proliferating and migrating potentials of ESCC cells (Figure 1).Disruption of endogenous cyclin B1 protein via siRNA knockdowntechnique was shown to substantially suppress cell proliferation, col-ony formation and invasion (Figure 3), suggesting an association ofcyclin B1 with tumor cell invasive growth in vitro. Additionally, theresults generated from mouse models indicate overexpression of cy-clin B1, promotes tumor invasive growth in vivo and probably resultsin lung metastasis. Strikingly, disruption of endogenous cyclin B1 inhighly metastatic EC9706 cells by siRNA technique has shown toblock their ability in lung metastasis (Table I). Therefore, these find-ings have demonstrated that cyclin B1 plays a key role in ESCCmetastasis from the primary tumor to lung and, at least in part, elu-cidate the essence of overexpression of cyclin B1 inducing the pooroutcome in ESCC patients.

The underlying mechanisms of cyclin B1 contributing to the ma-lignancies (invasion and metastasis) of esophageal cancer might in-volve several aspects. It is well known that cell cycle checkpoints playan important role in the control mechanisms that ensure the properexecution of cell cycle events. One of these checkpoints, the G2/Mcheckpoint, which is mainly related to cyclin B1/Cdc2, blocks entryinto mitosis when DNA is damaged (40). However, overexpression ofcyclin B1 may cause constitutive activation of cyclin B1-associatedCdc2 kinase and thus abrogate cell cycle G2/M arrest (38). As a con-sequence, genomic instability may occur due to the uncontrolled mi-totic activity, which would be able to establish the malignant potentialof the ESCC. The restructuring of the cancer cell genome, whichpermits tumor cells to overcome the normal structures against exces-sive multiplication and metastasis, may thus be due to abnormal cell

cycle control (41). So disruption of control at G2/M transition, espe-cially at G2/M checkpoint, caused by overexpression of cyclin B1could lead to genomic instability during tumor cell evolution.

Tumor metastasis is a complex multistep process that involves thedetachment of cancer cells from the primary tumor mass, intravasa-tion, extravasation and the establishment of new foci in a remoteorgan (2,42). One of the key steps in cancer metastasis is extravasa-tion. In the present study, we employed TEM assay and found thatoverexpression of cyclin B1 offered cells a stronger ability in trans-migrating through the ECEC and HLEC. TEM is a dynamic processthat involves the constant breaking and remaking of intercellular con-tacts and is accompanied by drastic changes in cell shape and cyto-skeletal reorganization in both the tumor cell and its interactingendothelial cells (43,44). Extravasation of cancer cells is a compli-cated event that involves both adhesive interactions and chemotaxis(43,45). Activated cyclin B1/Cdc2 kinase is responsible for initiatingprofound changes in the cellular architecture. Cyclin B1/Cdc2 phos-phorylates the nuclear lamins (46) and vimentin (47), to reorganizethe karyoskeletal and cytoskeletal intermediate filament networks,respectively, and promotes microfilament reorganization by phos-phorylating caldesmon (48). Cyclin B1/Cdc2 kinase activity alsochanges the nucleating ability of centrosomes and the dynamics ofmicrotubule polymerization (49). Most probably, overexpression ofcyclin B1 may result in cytoskeletal change and then promotes ex-travasation of ESCC. This is in agreement with the observation thatcyclin B1 enhances cell migration and invasive growth (Figure 2).

Among the many changes in gene expression and protein functionthat occur during tumor progression, alterations in cell–cell and cell–matrix adhesion seem to have a central role in facilitating tumor cellmigration, invasion and metastasis (4). Along with the enhanced

Fig. 4. Overexpression of cyclin B1 facilitates TEM of KYSE150. (A) KYSE150, pcDNA3.1 and two high-cyclin B1 cell lines (stained red with PKH26)transmigration through ECEC. (B) The transmigration cell number was counted in 10 random fields. (C) KYSE150, pcDNA3.1 and two high-cyclin B1 cell lines(stained red with PKH26) transendothelial through HLEC. (D) The transendothelial cell number was counted in 10 random fields. (E) Tumor cell–ECEC adhesionassay. (F) Tumor cell–HLEC adhesion assay.

Y.Song et al.

312



ownloaded from


Fig. 5. Overexpression of cyclin B1 induces alterations of EMT-related molecules in KYSE150 cell lines and inhibited cyclin B1 expression by siRNA-transfectedresults in E-cadherin revert in three ESCC cell lines. (A) Expression of epithelial proteins, including E-cadherin and b-catenin, and mesenchymal proteins,including fibronectin and N-cadherin, was examined by immunoblotting in pcDNA3.1 and one high-CycB1 cells. NF-jB and Twist were also detected. Three celllines (KYSE450, KYSE180 and EC9706) were transfected with either cyclin B1 siRNA or control siRNA and detected expression of several proteins, such ascyclin B1, E-cadherin and Twist. b-Actin is used as a loading control. (B) Immunofluorescence staining of E-cadherin, b-catenin, fibronectin, cyclin B1, NF-jBand N-cadherin analyzed by confocal microscopy in pcDNA3.1 and one high-CycB1 cells. The green and red signals represent the staining of correspondingprotein and the blue signal represents the nuclear DNA staining by 4#,6-diamidino-2-phenylindole.


313



ownloaded from


metastatic capability, overexpression of cyclin B1 causes suppressedexpression of E-cadherin and increased expression of N-cadherin.Interestingly, we also observed a direct correlation between the pro-tein expression level of cyclin B1 and NF-jB in invasive ESCC celllines. This finding reflects the involvement of NF-jB-mediatedpathway in cyclin B1-induced tumor malignancy and further suggeststhat cyclin B1-induced EMT by NF-jB-dependent pathway may bean important mechanism for tumor metastasis. Indeed, it has beenfound that NF-jB activity is necessary for cells to be maintained ina mesenchymal transition in a model of breast cancer progression(50–52).

The current EMT models are derived from different tissues anddiversities of molecular mechanisms contribute to the plasticity ofepithelial cells. In the majority of epithelial cell types tested and thetransgenic mouse tumor models as well, transforming growth factor-b(TGF-b) signaling is found to cooperate with the oncogenic Ras orreceptor tyrosine kinases to cause EMT and metastasis (53–57). Nev-ertheless, TGF-b reduces cell proliferation through transcriptionalsuppression of the cyclins, including cyclin D1 (58) and throughtranscriptional upregulation of cyclin-dependent kinase inhibitors.Decreases in cell proliferation have been tied to TGF-b-mediatedincreases in both p21 (59–63) and p27 (64–66).

In a variety of different cell types, TGF-b inhibits cell proliferationby its ability to downregulate the protooncogene c-myc and to inducethe transcription of cyclin-dependent kinase inhibitors such asp15INK4b, p27Kip1 and/or p21Cip1 (67).

Interestingly, TGF-b both represses inhibitor of DNA-binding pro-teins (68,69) and activates Snail family members (70), thus establish-ing direct links between TGF-b signaling, E-cadherin repression andEMT initiation. Additionally, our observations further demonstratea direct correlation between the protein expression level of cyclinB1 and Twist, which may play an essential role in tumor metastasisby inducing an epithelial–mesenchymal-like transition (11). However,in spite of repeated attempts, we could not observe an induction ofSnail in our cell lines that have undergone an EMT. Interestingly, theE-cadherin repressor Twist has been suggested as possible down-stream targets of NF-jB (50). These events appear to elucidate thatTwist might play an important role in cyclin B1-induced EMT inESCC cell lines via NF-jB signaling pathway. Hence, the facts thatboth TGFb as a cell cycle inhibitor and cyclin B1 as a cell cycleenhancer are able to induce EMT suggest that many such EMT-inducing transcription factors may be exploited opportunistically bydifferent types of tumor malignancies and metastatic powers (11).

In our study, we observed that ESCC xenograft expressing highlevel of cyclin B1 exhibited significant incidences of metastases inlungs as compared with its parental cells (Table I). Interestingly, wehave found a significant correlation between expression of cyclin B1and lung metastasis in nude mice, which is in agreement with cyclinB1-induced EMT in ESCC cell lines in vitro. This observation isconsistent with the previous findings that squamous cell carcinomasof head and neck xenografts in mouse model generated from an squa-mous cell corcinomas of head and neck cell line demonstrate thatsquamous cell corcinomas of head and neck cells-gained EMT fea-tures are more powerful in metastasis to lung (71).

The cell cycle progression is tightly regulated by a series of delicatecontrols that act on the transcription of cyclin genes, the degradationof cyclin proteins, and the modification of the cell cycle-related kinase(72,73). Multiple positive and negative feedback loops greatly con-tribute to the control of cell cycle progression (72). Cell-fate acqui-sition would depend on cell cycle progression either by a mechanismthat control cell division or by a mechanism that is intrinsic to the cellcycle machinery (74). Cell-fate diversity is, in consequence, related tothe cell cycle by cell proliferation and cell differentiation. Multitudegenetic changes occur during the evolution of the normal cells intocancer cells. Specific genetic changes can abrogate the fidelityachieved by the coordinated activity of cyclin-dependent kinase,checkpoint controls and repair pathways.

In our study, cyclin B1 that plays an important role as a mitoticcyclin in the G2/M phase transition during the cell cycle, and over-

expression of cyclin B1 promotes ESCC cell lines invasive growth andmetastasis by inducing EMT. Cancer develops when molecular path-ways that control the delicate balance between proliferation, differ-entiation and apoptosis undergo genetic deregulation.

In summary, the characterization of the mechanisms by which cy-clin B1 contributes to the invasion and metastasis of ESCC is ofimportance. In addition to well understanding the biological principleof ESCC malignant progression, it may provide significant clinicalapplication, including identifying cyclin B1 as a molecular marker inESCC early diagnosis and an indicator of prognosis. Furthermore,cyclin B1 would serve as an important molecular target for drugdiscovery, which should lead to new therapeutic approaches for anti-metastatic cancer treatments and exploiting the clinical practice ontumor-specific diagnosis and treatment.

Funding

The 973 National Key Fundamental Research Program of China(2002 CB513101); National Natural Science Foundation of China(30225018, 30721001).

Acknowledgements

Conflict of Interest Statement: None declared.

References

1.Chambers,A.F. et al. (2002) Dissemination and growth of cancer cells in

metastatic sites. Nat. Rev. Cancer, 2, 563–572.2.Fidler,I.J. (2003) The pathogenesis of cancer metastasis: the ‘seed and soil’

hypothesis revisited. Nat. Rev. Cancer, 3, 453–458.3.Karreth,F. et al. (2004) Twist induces an epithelial-mesenchymal transition

to facilitate tumor metastasis. Cancer Biol. Ther., 3, 1058–1059.4.Christofori,G. (2006) New signals from the invasive front. Nature, 441,

444–450.5.Radisky,D.C. et al. (2005) Rac1b and reactive oxygen species mediate

MMP-3-induced EMT and genomic instability. Nature, 436, 123–127.6.Boire,A. et al. (2005) PAR1 is a matrix metalloprotease-1 receptor that

promotes invasion and tumorigenesis of breast cancer cells. Cell, 120,

303–313.7.Minn,A.J. et al. (2005) Genes that mediate breast cancer metastasis to lung.

Nature, 436, 518–524.8.Clark,E.A. et al. (2000) Genomic analysis of metastasis reveals an essential

role for RhoC. Nature, 406, 532–535.9.Kang,Y. et al. (2003) A multigenic program mediating breast cancer me-

tastasis to bone. Cancer Cell, 3, 537–549.10.Bao,S. et al. (2004) Periostin potently promotes metastatic growth of colon

cancer by augmenting cell survival via the Akt/PKB pathway. Cancer Cell,

5, 329–339.11.Yang,J. et al. (2004) Twist, a master regulator of morphogenesis, plays an

essential role in tumor metastasis. Cell, 117, 927–939.12.Parkin,D.M. et al. (1988) Estimates of the worldwide frequency of sixteen

major cancers in 1980. Int. J. Cancer, 41, 184–197.13.Tew,W.P. et al. (2005) Targeted therapies for esophageal cancer. Oncolo-

gist, 10, 590–601.14.Hunter,T. et al. (1991) Cyclins and cancer. Cell, 66, 1071–1074.15.Peters,J.M. (2002) The anaphase-promoting complex: proteolysis in mito-

sis and beyond. Mol. Cell, 9, 931–943.16.Soria,J.C. et al. (2000) Overexpression of cyclin B1 in early-stage non-

small cell lung cancer and its clinical implication. Cancer Res., 60, 4000–

4004.17.Banerjee,S.K. et al. (2000) Expression of cdc2 and cyclin B1 in Helico-

bacter pylori-associated gastric MALT and MALT lymphoma: relationship

to cell death, proliferation and transformation. Am. J. Pathol., 156, 217–

225.18.Allan,K. et al. (2000) Overexpression of cyclin A and cyclin B1 proteins in

astrocytomas. Arch. Pathol. Lab. Med., 124, 216–220.19.Hiroshi,M. et al. (1999) Determination of the prognostic significance of

cyclin B1 overexpression in patients with esophageal squamous cell carci-

noma. Virchows Archiv., 434, 153–158.

Y.Song et al.

314



ownloaded from


20.Kallakury,B.V.S. et al. (1999) The prognostic significance of proliferation-associated nucleolar protein p120 expression in prostate adenocarcinoma.Cancer, 85, 1569–1576.

21.Kushner,J. et al. (1999) Aberrant expression of cyclin A and cyclin B1proteins in oral carcinoma. J. Oral Pathol. Med., 28, 77–81.

22.Kawamoto,H. et al. (1997) Expression of the G2-M checkpoint regulatorscyclin B1 and cdc2 in nonmalignant and malignant human breast lesions:immunocytochemical and quantitative image analyses. Am. J. Pathol., 150,15–23.

23.Dutta,A. et al. (1995) Cyclins as markers of tumor proliferation: immuno-cytochemical studies in breast cancer. Proc. Natl Acad. Sci. USA, 92, 5386–5390.

24.King,J. et al. (2004) Structural and functional characteristics of lungmacro- and microvascular endothelial cell phenotypes. Microvasc. Res.,67, 139–151.

25.Zhong,X. et al. (2004) Construction of human liver cancer vascular endo-thelium cDNA expression library and screening of the endothelium-associated antigen genes. World J. Gastroenterol., 10, 1402–1408.

26.Salmon,P. et al. (2000) Reversible immortalization of human primarycells by lentivector-mediated transfer of specific genes. Mol. Ther., 2,404–414.

27.Yuan,J. et al. (2004) Cyclin B1 depletion inhibits proliferation and inducesapoptosis in human tumor cells. Oncogene, 23, 5843–5852.

28.Li,W. et al. (2005) Overexpression of stefin A in human esophageal squa-mous cell carcinoma cells inhibits tumor cell growth, angiogenesis, inva-sion, and metastasis. Clin. Cancer Res., 11, 8753–8762.

29.Lee,T.H. et al. (2003) Vascular endothelial growth factor modulates thetransendothelial migration of MDA-MB-231 breast cancer cells throughregulation of brain microvascular endothelial cell permeability. J. Biol.Chem., 278, 5277–5284.

30.Woodward,J.K. et al. (2002) An in vitro assay to assess uveal melanomainvasion across endothelial and basement membrane barriers. Invest. Oph-thalmol. Vis. Sci., 43, 1708–1714.

31.Bild,T. et al. (2004) Discovery of inhibitors of MCF-7 tumor cell adhesionto endothelial cells and investigation on their mode of action. Arch. Pharm.(Weinheim), 337, 687–694.

32.Li,Y.M. et al. (2004) Upregulation of CXCR4 is essential for HER2-mediated tumor metastasis. Cancer Cell, 6, 459–469.

33.Nozoe,T. et al. (2002) Significance of cyclin B1 expression as an indepen-dent prognostic indicator of patients with squamous cell carcinoma of theesophagus. Clin. Cancer Res., 8, 817–822.

34.Qi,J. et al. (2006) Involvement of Src family kinases in N-cadherin phos-phorylation and beta-catenin dissociation during transendothelial migrationof melanoma cells. Mol. Biol. Cell, 17, 1261–1272.

35.Thiery,J.P. (2002) Epithelial-mesenchymal transitions in tumour progres-sion. Nat. Rev. Cancer, 2, 442–454.

36.Thiery,J.P. (2003) Epithelial-mesenchymal transitions in development andpathologies. Curr. Opin. Cell Biol., 15, 740–746.

37.Boyer,B. et al. (1993) Epithelium-mesenchyme interconversion as exampleof epithelial plasticity. APMIS, 101, 257–268.

38.Park,M. et al. (2000) Constitutive activation of cyclin B1-associated cdc2kinase overrides p53-mediated G2-M arrest. Cancer Res., 60, 542–545.

39.Guardavaccaro,D. et al. (2006) Stabilizers and destabilizers controlling cellcycle oscillators. Mol. Cell, 22, 1–4.

40.Taylor,W.R. et al. (2001) Regulation of the G2/M transition by p53.Oncogene, 20, 1803–1815.

41.Yu,M. et al. (2002) Immune recognition of cyclin B1 as a tumor antigen isa result of its overexpression in human tumors that is caused by non-functional p53. Mol. Immunol., 38, 981–987.

42.Pantel,K. et al. (2004) Dissecting the metastatic cascade. Nat. Rev. Cancer,4, 448–456.

43.Voura,E.B. et al. (2001) Involvement of integrin alpha(v)beta(3) and celladhesion molecule L1 in transendothelial migration of melanoma cells.Mol. Biol. Cell, 12, 2699–2710.

44.Brandt,B.H. et al. (1999) c-erbB-2/EGFR as dominant heterodimerizationpartners determine a motogenic phenotype in human breast cancer cells.FASEB J., 13, 1939–1949.

45.Ramjeesingh,R. et al. (2003) Interleukin-8 secreted by endothelial cellsinduces chemotaxis of melanoma cells through the chemokine receptorCXCR1. FASEB J., 17, 1292–1294.

46.Peter,M. et al. (1991) Disassembly of in vitro formed lamin head-to-tailpolymers by CDC2 kinase. EMBO J., 10, 1535–1544.

47.Chou,Y.H. et al. (1990) Intermediate filament reorganization during mitosisis mediated by p34cdc2 phosphorylation of vimentin. Cell, 62, 1063–1071.

48.Yamashiro,S. et al. (1991) Phosphorylation of non-muscle caldesmon byp34cdc2 kinase during mitosis. Nature, 349, 169–172.

49.Verde,F. et al. (1992) Control of microtubule dynamics and length by cyclinA- and cyclin B-dependent kinases in Xenopus egg extracts. J. Cell Biol.,118, 1097–1108.

50.Kang,Y. et al. (2004) Epithelial-mesenchymal transitions: twist in devel-opment and metastasis. Cell, 118, 277–279.

51.Huber,M.A. et al. (2004) NF-{kappa}B is essential for epithelial-mesenchymal transition and metastasis in a model of breast cancer pro-gression. J. Clin. Invest., 114, 569–581.

52.Bachelder,R.E. et al. (2005) Glycogen synthase kinase-3 is an endogenousinhibitor of Snail transcription: implications for the epithelial-mesenchymaltransition. J. Cell Biol., 168, 29–33.

53.Grunert,S. et al. (2003) Diverse cellular and molecular mechanisms con-tribute to epithelial plasticity and metastasis. Nat. Rev. Mol. Cell Biol., 4,657–665.

54.Xie,L. et al. (2004) Activation of the Erk pathway is required for TGF-beta1-induced EMT in vitro. Neoplasia, 6, 603–610.

55.Seton-Rogers,S.E. et al. (2004) Cooperation of the ErbB2 receptor andtransforming growth factor beta in induction of migration and invasion inmammary epithelial cells. Proc. Natl Acad. Sci. USA, 101, 1257–1262.

56.Ueda,Y. et al. (2004) Overexpression of HER2 (erbB2) in human breastepithelial cells unmasks transforming growth factor beta-induced cell mo-tility. J. Biol. Chem., 279, 24505–24513.

57.Siegel,P.M. et al. (2003) Transforming growth factor beta signaling impairsNeu-induced mammary tumorigenesis while promoting pulmonary metas-tasis. Proc. Natl Acad. Sci. USA, 100, 8430–8435.

58.Ko,T.C. et al. (1998) TGF-beta1 effects on proliferation of rat intestinalepithelial cells are due to inhibition of cyclin D1 expression. Oncogene, 16,3445–3454.

59.Elbendary,A. et al. (1994) Transforming growth factor beta 1 can induceCIP1/WAF1 expression independent of the p53 pathway in ovarian cancercells. Cell Growth Differ., 5, 1301–1307.

60.Datto,M.B. et al. (1995) Transforming growth factor beta induces thecyclin-dependent kinase inhibitor p21 through a p53-independent mecha-nism. Proc. Natl Acad. Sci. USA, 92, 5545–5549.

61.Li,C.Y. et al. (1995) Potential role of WAF1/Cip1/p21 as a mediator ofTGF-beta cytoinhibitory effect. J. Biol. Chem., 270, 4971–4974.

62.Yoo,Y.D. et al. (1999) TGF-beta-induced cell-cycle arrest through thep21(WAF1/CIP1)-G1 cyclin/Cdks-p130 pathway in gastric-carcinomacells. Int. J. Cancer, 83, 512–517.

63.Wolfraim,L.A. et al. (2004) p21Cip1 and p27Kip1 act in synergy to alterthe sensitivity of naive T cells to TGF-beta-mediated G1 arrest throughmodulation of IL-2 responsiveness. J. Immunol., 173, 3093–3102.

64.Polyak,K. et al. (1994) p27Kip1, a cyclin-Cdk inhibitor, links transforminggrowth factor-beta and contact inhibition to cell cycle arrest. Genes Dev., 8,9–22.

65.Bouchard,C. et al. (1997) Effect of TGF-beta1 on cell cycle regulatoryproteins in LPS-stimulated normal mouse B lymphocytes. J. Immunol.,159, 4155–4164.

66.Kamesaki,H. et al. (1998) TGF-beta 1 induces the cyclin-dependent kinaseinhibitor p27Kip1 mRNA and protein in murine B cells. J. Immunol., 160,770–777.

67.Massague,J. et al. (2000) TGFbeta signaling in growth control, cancer, andheritable disorders. Cell, 103, 295–309.

68.Kondo,M. et al. (2004) A role for Id in the regulation of TGF-beta-inducedepithelial-mesenchymal transdifferentiation.Cell Death Differ., 11, 1092–1101.

69.Kowanetz,M. et al. (2004) Id2 and Id3 define the potency of cell prolifer-ation and differentiation responses to transforming growth factor beta andbone morphogenetic protein. Mol. Cell. Biol., 24, 4241–4254.

70.Peinado,H. et al. (2003) Transforming growth factor beta-1 induces snailtranscription factor in epithelial cell lines: mechanisms for epithelial mes-enchymal transitions. J. Biol. Chem., 278, 21113–21123.

71.Zhang,X. et al. (2006) Understanding metastatic SCCHN cells from uniquegenotypes to phenotypes with the aid of an animal model and DNA micro-array analysis. Clin. Exp. Metastasis, 23, 209–222.

72.Nasmyth,K. (1993) Control of the yeast cell cycle by the Cdc28 proteinkinase. Curr. Opin. Cell Biol., 5, 166–179.

73.Nurse,P. (1990) Universal control mechanism regulating onset of M-phase.Nature, 344, 503–508.

74.Fichelson,P. et al. (2005) Cell cycle and cell-fate determination inDrosophila neural cell lineages. Trends Genet., 21, 413–420.

Received May 30, 2007; revised October 26, 2007;accepted November 20, 2007


315



ownloaded from


Documents

OverexpressionofcyclinB1inhumanesophagealsquamouscellcarci ... · Adhesion assay Tumor cell–endothelial cell adhesion assay was performed as described (31,32). In total, 96-well