Available online at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/yexcr

E X P E R I M E N T A L C E L L R E S E A R C H 3 1 8 ( 2 0 1 2 ) 1 7 0 7 – 1 7 1 5

0014-4827/$ - see frohttp://dx.doi.org/10.

nCorresponding au

Luoyu Road, Wuhan

E-mail address:

Research Article

Tumor necrosis factor-a enhanced fusions between oral

squamous cell carcinoma cells and endothelial cells via

VCAM-1/VLA-4 pathway

Kai Songa, Fei Zhua, Han-zhong Zhanga, Zheng-jun Shanga,b,n

aThe State Key Laboratory Breeding Base of Basic Science of Stomatology (Hubei-MOST), Key Laboratory for Oral Biomedicine Ministry

of Education, Wuhan University, Wuhan, ChinabFirst Department of Oral and Maxillofacial Surgery, School and Hospital of Stomatology, Wuhan University, Wuhan, China

a r t i c l e i n f o r m a t i o n

Article Chronology:

Received 10 February 2012

Received in revised form

1 May 2012

Accepted 24 May 2012

Available online 1 June 2012

Keywords:

Oral squamous carcinoma

VCAM-1

VLA-4

Cell fusion

Tumor necrosis factor-a

nt matter & 2012 Elsevier1016/j.yexcr.2012.05.022

thor at: First Department

, 430079, China. Fax: þ86

shangzhengjun@hotmail.co

a b s t r a c t

Fusion between cancer cells and host cells, including endothelial cells, may strongly modulate

the biological behavior of tumors. However, no one is sure about the driving factors and

underlying mechanism involved in such fusion. We hypothesized in this study that inflamma-

tion, one of the main characteristics in tumor microenvironment, serves as a prominent catalyst

for fusion events. Our results showed that oral cancer cells can fuse spontaneously with

endothelial cells in co-culture and inflammatory cytokine tumor necrosis factor-a (TNF-a)

increased fusion of human umbilical vein endothelium cells and oral cancer cells by up to 3-fold

in vitro. Additionally, human oral squamous cell carcinoma cell lines and 35 out of 50 (70%) oral

squamous carcinoma specimens express VLA-4, an integrin, previously implicated in fusions

between human peripheral blood CD34-positive cells and murine cardiomyocytes. Expression of

VCAM-1, a ligand for VLA-4, was evident on vascular endothelium of oral squamous cell

carcinoma. Moreover, immunocytochemistry and flow cytometry analysis revealed that

expression of VCAM-1 increased obviously in TNF-a-stimulated endothelial cells. Anti-VLA-4

or anti-VCAM-1 treatment can decrease significantly cancer–endothelial adhesion and block

such fusion. Collectively, our results suggested that TNF-a could enhance cancer–endothelial

cell adhesion and fusion through VCAM-1/VLA-4 pathway. This study provides insights into

regulatory mechanism of cancer–endothelial cell fusion, and has important implications for the

development of novel therapeutic strategies for prevention of metastasis.

& 2012 Elsevier Inc. All rights reserved.

Introduction

Interaction of cancer cells with endothelial cells is a pivotal step

in both tumor angiogenesis and metastatic dissemination [1,2].

Recently, Mortensen et al. have observed spontaneous fusion

Inc. All rights reserved.

of Oral and Maxillofacial S

27 87873260.

m (Z.-j. Shang).

between cancer cells and endothelial cells in vitro and in vivo,

which provide a novel type of tumor–endothelial cell interaction

and may strongly modulate the biological behavior of tumors [3].

Therefore, understanding the underlying mechanisms is crucial

for the development of metastasis-targeting manipulations.

urgery, School and Hospital of Stomatology, Wuhan University, 237

E X P E R I M E N T A L C E L L R E S E A R C H 3 1 8 ( 2 0 1 2 ) 1 7 0 7 – 1 7 1 51708

As heterotypic cell fusion is extremely rare under normal

conditions, it is possible that pre-conditioning of organs by

injury would increase the frequency of fusion [4]. Indeed,

heterokaryon formation increased markedly under irradiation.

Interestingly, Nygren et al. found that such irradiation-induced

fusion could be inhibited by using the anti-inflammatory drug

prednisolone [5]. Therefore, inflammation seems to serve as a

prominent catalyst for fusion events. Tumor necrosis factor

(TNF)-a is a key inflammatory cytokine which was first identified

for its ability of anti-tumor therapy [6,7]. Recently, emerging

evidences have shown TNF-a was a major mediator of cancer-

related inflammation and acted as a tumor-promoting factor in

microenvironment [8–10]. It has been demonstrated that TNF-awas involved in many aspects of tumor development and

progression, including tumor initiation, tumor cell proliferation

and survival, tumor angiogenesis, formation of invasive pheno-

type, and epithelial–mesenchymal transition (EMT) [8,9,11].

However, it has been not fully defined regarding whether and

how tumor cell-produced TNF-a mediates tumor–endothelial

cell interaction, especially cell fusion, to facilitate tumor growth

and metastasis.

Vascular cell adhesion molecule-1 (VCAM-1) is a cell surface

receptor expressed on activated endothelial and mesothelial cells.

Through its interaction with very late activation antigen

4 (VLA-4) expressed on leukocytes, VCAM-1 functions to regulate

leukocyte attachment and extravasation across endothelial and

mesothelial monolayers at sites of inflammation [12–15]. Recent

studies have shown that the increased expression of VCAM-1 in

inflammatory tumor microenvironment was correlated with devel-

opment, progression of several types of malignant tumors, and VLA-4

(a4b1) was expressed in a variety of cancers [16–19]. In addition, it

has been reported that inflammatory cytokine interleukin-1 (IL-1)

promoted tumor cells adhesion to cultured human endothelial cells

in vitro and augmented experimental lung metastases through

VCAM-1/VLA-4 signaling [2,20]. As the attachment and adhesion is

a prerequisite for tumor–endothelial cell fusion, these findings raised

the possibility that VCAM-1/VLA-4 pathway might be also involved

in cancer–endothelial cell adhesion and thereafter fusion in TNF-a-

induced tumor inflammatory microenvironment.

In present study, we investigated the effects of TNF-a on oral

cancer–endothelial cell fusion and adhesion, and its link with

VCAM-1/VLA-4 pathway. Here, we demonstrated that TNF-astimulation enhanced OSCC–endothelial cell adhesion and fusion

as well as VCAM-1 expression on endothelial cells in vitro. We

also confirmed VLA-4 expression on human oral squamous cell

carcinoma (OSCC) for the first time. Pretreatment of cells with

anti-VLA-4 or anti-VCAM-1 almost abolished the TNF-a-

enhanced OSCC–endothelial cell adhesion and fusion. Taken

together, this study provides important insights into how cancer

cells interact directly with vascular endothelial cells in tumor

microenvironment, and points to a potential mechanism linking

inflammation and cancer metastasis.

Materials and methods

Cell lines and cell culture

The human OSCC cell lines, CAL-27 (CRL-2095, ATCC) and SCC-4

(CRL-1624, ATCC), were kindly provided by Affiliated Ninth

People’s Hospital, Shanghai Jiaotong University, Shanghai

Research Institute of Stomatology, Shanghai, China. The cells

were separately cultured in DMEM and DMEM/F12 (Hyclone, UT,

USA) supplemented with 10% FBS (Gibco, Carlsbad, Calif, USA).

Human umbilical vein endothelial cells (HUVECs) were obtained

from the China Center for Type Culture Collection and were

cultured in RPMI-1640 medium (Hyclone, UT, USA) containing

10% FBS. The human osteosarcoma cell line, MG-63 was provided

by the Key Laboratory for Oral Biomedical Engineering of

Ministry of Education at Wuhan University, China and the

osteosarcoma cell line was cultured in DMEM supplemented

with 10% fetal bovine serum. All cells were cultured at 37 1C in an

atmosphere containing 5% CO2.

Immunohistochemistry

Surgical specimens from 50 patients with oral squamous cell

carcinoma were fixed in 10% buffered formalin and embedded in

paraffin. Deparaffinized sections underwent antigen retrieval by

microwaving and were blocked with 10% goat serum, following

incubation with antibody to human VLA-4 (1:600; epitomics) or

VCAM-1 (1:50; Sigma) and the presence of which was detected

with biotin-conjugated secondary antibody and avidin-peroxi-

dase. Nuclei were counterstained with haematoxylin.

Cell fusion assays

PHK26 labeled HUVECs were co-cultured with CFSE labeled CAL-

27 (2�105 cells in total) for 24 h, and then were detected using a

fluorescence microscope (Leica, Wetzlar, Germany) and analyzed

by flow cytometry (Becton Dickenson, Heidelberg, Germany).

Orange-like cells were counted in 3 random fields in a 100x

magnification and each experiment was repeated at least twice.

Adhesion assay

HUVECs were seeded into 12 of 24 wells and starved for 12 h

with serum-free medium and then stimulated with TNF-a(10 ng/ml; Peprotech) for another 24 h in a medium containing

2% FBS. CAL-27 cells (1.2�106) were incubated in 5 mg/ml of the

fluorescent dye CFSE (Sigma) at 37 1C for 15 min. HUVECs were

washed three times with DMEM before the dye-loaded cells

(1�105 cells/well) were added and incubated at 37 1C. After

60 min, cell suspensions were withdrawn, and the HUVECs were

gently washed with DMEM. Images were obtained using a

fluorescence microscope with a camera (Leica, Wetzlar,

Germany). Analyses were repeated three times and the results

are the mean values of the three independent experiments.

Block assay

Anti-VLA-4 mAb (60 mg/ml; Millipore) or anti-VCAM-1 mAb

(50mg/ml; Biolegend) was used to block VLA-4 or VCAM-1

functionally, and cells were incubated with mAb for 1 h at

37 1C before the start of the assay. TNF-a stimulated endothelial

cells and CAL-27 cells exposed to anti-VCAM-1 antibody or anti-

VLA-4 antibody for 1 h, and then were co-cultured for another

24 h at 37 1C in a humidified 5% CO2 atmosphere.

E X P E R I M E N T A L C E L L R E S E A R C H 3 1 8 ( 2 0 1 2 ) 1 7 0 7 – 1 7 1 5 1709

Immunocytochemical analysis

Endothelial cells growing on glass slides were incubated with

TNF-a (10 ng/ml) for 24 h. The cells were washed with PBS, fixed

with 3.7% paraformaldehyde in phosphate-buffered saline for

10 min and air dried. Following three further washes with PBS,

the cells were incubated with 10% (v/v) goat serum for 20 min to

suppress nonspecific binding of IgG. After a further wash with

PBS, cells were incubated with anti-human VCAM-1 (1:50;

Sigma) in PBS with 2.5% BSA for 1 h (Molecular Probes).

Flow cytometry analysis

The cells were removed by 0.05% trypsin/0.02% EDTA treatment,

resuspended in 0.5% BSA/PBS and then distributed 100 ml/tube of

cell suspension (1.0�106 cells/tube), following incubation in

suspension for 45 min at 4 1C with PE-conjugated anti-VCAM-1

or PE-conjugated isotypic IgG negative control. After washing

twice in cold PBS, the cells were fixed in 1% paraformaldehyde as

a single-cell suspension for analysis by a fluorescence-activated

cell sorter (Becton Dickenson, Heidelberg, Germany).

Western blot analysis

Aliquots of 20 mg of protein were subjected to 10% sodium

dodecyl sulfate–polyacrylamide gel electrophoresis for 1 h

30 min at 110 V. The separated proteins were transferred to

nitrocellulose membranes for 2 h at 200 mA. The membranes

were blocked with 5% non-fat milk in Tris-buffered saline

containing 0.05% Tween 20 for 1 h at room temperature. Then,

the membranes were incubated with anti-VLA-4 antibody

Fig. 1 – Spontaneous cell fusion between squamous cancer and en

(membrane labeling) before being co-cultured with CAL-27 cells

cells showed the cytoplasm and cell nuclear dye. (c) A hybrid cell c

incorporated bi-fluorescent signals and contained three nuclei. Sc

this figure legend, the reader is referred to the web version of th

(1:1500; eptitomic) or anti-VCAM-1 antibody (1:3000; epitomic)

overnight at 4 1C, and the bound antibody was detected with

horseradish peroxidase-conjugated, anti-rabbit IgG (Pierce Che-

mical, Rockford, IL, USA). Western blot experiments were

repeated at least twice to confirm the results.

Statistical analysis

Statistical analysis was performed using the Student’s t-test,

p valueso0.05 were considered significant.

Results

Spontaneous oral squamous cell carcinoma–endothelialcell fusions

To investigate whether oral squamous cell carcinoma cells can

fuse with endothelial cells spontaneously, CFSE-labeled CAL-27

cells were co-cultured with HUVECs labeled with the red fluor-

escent probe PKH26 at a 1:1 rate for 24 h without any fusogenic

reagents. After 24 h of co-culture, cells that incorporated red and

green fluorescent signals and contained two or more nuclei were

considered ECs that had fused with cancer cells. Fig. 1 showed

that a CFSE/PKH26 cell contained three nuclei (Fig. 1d).

TNF-a enhanced oral squamous cell carcinoma–endothelial

cell fusion

Inflammatory responses play critical roles at different stages of

tumor development, including initiation, promotion, malignant

dothelial cells. (a) Endothelial cells were labeled with PKH26

and showed red fluorescent signals. (b) CFSE-labeled CAL-27

ontained three nuclei (blue arrow). (d) Noted that a hybrid cell

ale bar, 20 lm. (For interpretation of the references to color in

is article.)

Fig. 2 – TNF-a treatment enhanced cell fusions. Endothelial cells were exposed to TNF-a conditions for 24 h before being

co-cultured with CAL-27 cells. TNF-a increased the percentage of bi-fluorescent cells, compared with control group. Scale bar,

50 lm.

E X P E R I M E N T A L C E L L R E S E A R C H 3 1 8 ( 2 0 1 2 ) 1 7 0 7 – 1 7 1 51710

conversion, invasion, and metastasis. To observe the fusion event

under TNF-a-induced inflammatory conditions, endothelial

cells were stimulated with TNF-a (10 ng/ml) for 24 h before

co-culturing with oral carcinoma cells at a 1:1 ratio for another

24 h. This dose of TNF-a has been widely used to investigate the

pro- inflammatory effects of TNF-a in cultured cells. The TNF-astimulation enhanced the fusion by up to 3-fold in vitro by

counting orange-like cells in 3 random fields in a 100x magni-

fication, compared with untreated cells (Fig. 2). FACS analysis

also showed that TNF-a stimulation increased the percentage of

bi-fluorescent cells to 1.6970.09% (Fig. 6a), compared with

untreated cell (0.370.02%) (Fig. 6b).

Expression of VLA-4/VCAM-1 in oral squamous cellcarcinoma

In this study, we investigated VLA-4 expression in surgical

specimens from 50 patients with oral squamous cell carcinoma.

Immunohistochemistry showed that 35 out of 50 human oral

squamous cancer specimens (70%) contained cancer cells that

reacted for VLA-4 in the cell membrane (Fig. 3Aa and Ab).

Additionally, we further tested the protein level expression in

two oral squamous line cells (CAL-27 and SCC-4). Immunoblot-

ting revealed a 150-kDa immunoreactive band in extracts of

CAL-27 cells (Fig. 3B). In deparaffinized sections from surgical

specimens, microvascular endothelial VCAM-1 expression was

assessed by immunohistochemistry. Immunohistochemistry

showed that expression of VCAM-1 was present in the cell

membrane and was evident in the microvessels of oral squamous

cell carcinoma (Fig. 3Ac and Ad).

TNF-a increased VCAM-1 expression on HUVECs

The close contact of cells is a prerequisite for cell–cell fusion, we

hypothesized that surface adhesion molecules play an important

role in the fusion event and the increase in cell fusion is

attributable to induction of adhesion molecules expression on

stimulated cells by TNF-a treatment. We found that VCAM-1

expression on stimulated endothelial cells was increased

following exposure to TNF-a treatment for 24 h by immuno-

cytochemistry and flow cytometry analysis (Fig. 4). Fig. 4

showed that VCAM-1 expression was upregulated on stimu-

lated endothelial cells in contrast to unstimulated cells. The

PE fluorescence profiles are shown in Fig. 4A. Basal VCAM-1

was negligible, giving a similar profile to that of isotype-

treated control cells. After 24 h exposure to TNF-a there was a

clear right shift in the fluorescence peak indicating upregula-

tion of cell surface VCAM-1 (Fig. 4A).

VCAM-1/VLA-4 mediated TNF-a-enhanced

cancer–endothelial cell adhesion and fusion

Cancer cells–EC interactions are regulated in part by the expres-

sion of specific adhesion molecules. Therefore, the effect of

VCAM-1 or VLA-4 on the adhesion of cancer cells–HUVECs was

investigated following exposure to inflammatory stimuli. To

examine that the induction of the VCAM-1 or adhesion molecule

pair is relevant to the adhesion and fusion process, we first

investigated the effect that adhesion of human CAL-27 to TNF-aactivated HUVECs. The result showed that the adhesion of

CAL-27 cells increased significantly when HUVECs were stimu-

lated with TNF-a (10 ng/ml) for 24 h (Fig. 5A). In contrast, HUVECs

and CAL-27 treated with anti-VCAM-1 antibody or anti-VLA-4

antibody for 1 h showed significant reduction of CAL-27 cells

adhering to ECs (Fig. 5Ca). We also found TNF-a enhanced fusion

could be blocked when cells were incubated with anti-VLA-4

antibody or anti-VCAM-1 antibody as opposed to control group

before co-culture. As shown in Fig. 5Cb, the frequency of fusion

occurrence was blocked by up to 70% in cells treated with the

anti-VLA-4 antibody, but not in cells treated with a BSA-matched

control. We also found that anti-VCAM-1 treatment blocked fusion

in a similar level (Fig. 5Cb). The quantitative data by FACS also

demonstrated that anti-VLA-4 and anti-VCAM-1 treatment inhib-

ited the frequency of bi-fluorescent cells to 0.6070.04% and

0.6470.06%, respectively (Fig. 6C and D).

Fig. 3 – VLA-4/VCAM-1 expression in squamous cell carcinoma. (Aa) Immunohistochemistry staining for VLA-4 in human

squamous cancer specimen. Note VLA-4 immunoreactivity (brown) in tumor cells. Magnification, x200. (Ab) Section to that

shown in Magnification x400. (Ac and Ad) Immunohistochemistry was used to detect VCAM-1 expression. Magnification, (Ac)

x200 and (Ad) x400. (B) Immunoblotting detects a 150-kDa band in extracts of squamous cancer cells. Immuoblotting detects an

immunoreactive band in extracts of CAL-27 cells (lane 3). No immunoreactive band was found in HUVECs (lane 2) and a

representive band was detected in extracts of MG-63 cells as a positive control (lane 4) [19]. (C) Immunoblotting detects VCAM-1

expression in the indicated time point. (For interpretation of the references to color in this figure legend, the reader is referred to

the web version of this article.)

E X P E R I M E N T A L C E L L R E S E A R C H 3 1 8 ( 2 0 1 2 ) 1 7 0 7 – 1 7 1 5 1711

Discussion

Interaction of cancer cells with endothelial cells plays an impor-

tant role in both tumor angiogenesis and metastasis [1,2]. Our

present study documented spontaneous fusion between cancer

cells and endothelial cells when human oral squamous cell

carcinoma cells were co-cultured with endothelial cells in vitro.

The fused cells observed in co-cultures expressed incorporated

red and green fluorescent signals and contained two or more

nuclei. However, such cells were never observed when HUVECs or

CAL-27 cells were cultured alone. These observations demon-

strate a new type of cancer–endothelial cell interaction that may

be of fundamental importance to the process of metastasis.

Firstly, tumor–endothelial cell fusion may contribute to increased

tumor angiogenesis because such fusions were usually seen in a

three-dimensional in vitro neovascularization model. Secondly,

tumor cells might acquire enhanced metastatic potential via

hybridization with cancer-related endothelial cells through per-

sistent expression of functions from both fusion partners [3]. In

fact, fusions induced to occur between macrophages and cancer

cells have been found to increase the metastatic capability of

cancer cells [21,22]. Cancer–normal cell fusion or cancer–cancer

cell fusion has been speculated to be related to cancer progression

and metastasis via phenotypic evolution. Recent findings by Lu

and Kang showed that the hybrids between bone- and lung-tropic

subline of the MDA-MB-231 had an important role in acquisition

of complex malignant traits and demonstrated the dual metas-

tasis organotropisms [23]. Metastasis is a multi-step process, and

one of pre-essential step is to survive and traffic in the circulation.

The fused cells may become more resistant to chemotherapeutic

drugs by acquiring drug resistance from both parental cells and

Fig. 4 – Surface expression of VCAM-1 in HUVECs. (A) Flow cytometric analysis of TNF-a-induced expression of cell-surface

VCAM-1. For FACS experiments, a PE-conjugated anti-VCAM-1 antibody binding to HUVEC suspensions was detected. Basal

VCAM-1 expression was negligible, but after 24 h exposure to TNF-a there was upregulation of cell-surface VCAM-1.

(B) Immunocytochemistry staining showed that TNF-a stimulated HUVECs induced an enhanced VCAM-1 surface expression

in comparison with unstimulated group. Scale bar, 50 lm.

E X P E R I M E N T A L C E L L R E S E A R C H 3 1 8 ( 2 0 1 2 ) 1 7 0 7 – 1 7 1 51712

generating new drug resistance. Recent research found that the

hybrids between 5-fluorouracil-resistant and methotrexate-resis-

tant mammary tumor cell lines became resistant to both drugs

and a new drug, melphalan [24]. Duelli and Lazebnik also

indicated that fusion of drug-sensitive transformed cells to

primary fibroblast cells produced hybrids that are resistant to

drug-induced apoptosis [25]. However, whether the hybrid cells

between cancer cells and endothelial cells were more metastatic

than original tumor cells or acquired drug resistance through cell

fusion were unclear. Finally, cancer–endothelial cell fusion might

be also a possible way for transendothelial migration of tumor

cells because cancer cells must interact with endothelium of

blood and lymphatic vessels and cross this barrier during cancer

metastasis.

The trigger and mechanisms underlying cancer–endothelial

fusion have been poorly understood because heterotypic cell

fusion is extremely rare, unpredictable and slow, and therefore,

difficult to study [4]. Inflammation is the main characteristic of

the tumor microenvironment, and plays multifaceted and critical

roles in every aspect of tumor development and progression as

well as the response to therapy [10]. Tumor cell entry into the

hepatic microvasculature can also trigger a rapid, proinflamma-

tory cascade that begins with increased local TNF-a and IL-1bproduction by activated Kupffer cells, suggesting that TNF-a is a

major inflammatory factor that acts as a master switch in

establishing an intricate link between inflammation and cancer

[9]. In the present study, we demonstrated for the first

time that inflammatory cytokine TNF-a was involved in oral

cancer–endothelial cell fusion in vitro. Under the stimulation of

TNF-a at a dose of 10 ng/ml, the frequency of oral cancer–

endothelial cell fusion increased by 3-fold in comparison with

normal conditions. One recent report demonstrated that chronic

inflammation resulting from severe dermatitis or autoimmune

encephalitis leads to extensive fusion of BMDCs with Purkinje

neurons [26]. Nygren et al. also found that irradiation-induced

fusion could be inhibited by using the anti-inflammatory drug

prednisolone [5]. Together with these findings, our results

provided further evidence to link inflammation with cancer,

and suggested that TNF-a may serve as a matchmaker for

heterotypic cell fusion, including cancer–endothelial cell fusion

in inflammatory tumor microenvironment.

Although NF-kB/Snail pathway has been involved in TNF-a-

induced tumor cell invasion and metastasis [27], there is nearly no

information available on how TNF-a promotes cancer–endothelial

cell fusion. Based on the recognition that the close contact and

adhesion of cells is a prerequisite for cell fusion, we hypothesized

that adhesion molecule VCAM-1/VLA-4 (a4b1 integrin) pathway

may mediate the TNF-a-enhanced oral cancer–endothelial cell fusion

in this study. With immunohistochemical and/or western blot

analysis, we found that VLA-4 expression was seen in 70% (35/50)

of human oral squamous cell carcinoma tissues as well as CAL-27

cells. The expression of VLA-4 on human oral squamous cancer may

contribute to tumorigenesis and have an underlying impact on

tumor metastasis. Rebhun et al. reported that B16-F1 tumors

metastasized to lymph nodes in 30% of mice, whereas B16a4þ

tumors generated lymph node metastases in 80% of mice. B16-F1

melanoma cells that were deficient in a4 integrins (B16a4�) were

nontumorigenic [28]. Collectively, these data show that the

a4 integrin expressed by cancer cells contributes to tumorigenesis

and may also facilitate metastasis to regional lymph nodes by

promoting stable adhesion of tumor cells to the lymphatic vascu-

lature. Then, we examine whether VCAM-1, a receptor of VLA-4,

express on endothelial cells and how endothelial activation changes

during the course of inflammatory response. Our data showed that

under TNF-a-induced inflammatory condition, stimulated endothe-

lial cells had an enhanced VCAM-1 expression in contrast to

Fig. 5 – Blocking assay analysis of the involvement of VCAM-1/VLA-4 in co-culture. (A) Adhesion of CAL-27 cells to (a) HUVECs and

to (b) TNF-a stimulated HUVECs. (c) Adhesion of CAL-27 cells to TNF-a stimulated HUVECs when blocking VLA-4 mAbs and

(d) blocking VCAM-1 mAbs. Original magnification, x200. (B) Fusion of CAL-27 cells with (a) TNF-a stimulated HUVECs when

blocking VLA-4 mAbs and (b) blocking VCAM-1 mAbs. Original magnification, x100. (C) Effect of blocking mAbs against VCAM-1

or VLA-4 during co-culture of CAL-27 cells with TNF-a stimulated HUVECs. (Ca) Blocking mAbs against VCAM-1 or VLA-4

treatment reduce strongly adhesion of CAL-27 to HUVECs compared with control group (mean7SD of 3 experiments).

���Po0.001 vs TNF-a (þ) group, ###Po0.001 vs TNF-a (þ) group. (Cb) Blocking mAbs against VCAM-1 or VLA-4 treatment reduce

significantly the percentage of bi-fluorescent cells in contrast with control group (mean7SD of 3 experiments). ��Po0.01 vs

TNF-a (þ) group, ##Po0.01 vs TNF-a (þ) group.

E X P E R I M E N T A L C E L L R E S E A R C H 3 1 8 ( 2 0 1 2 ) 1 7 0 7 – 1 7 1 5 1713

unstimulated endothelial cells. Moreover, we found that the expres-

sion of VCAM-1 was evident on tumor vascular endothelium. Such

expression changes under inflammatory microenvironment may be

related to migration of leukocyte extravasion and/or metastasis of

tumor [12,29–32]. It is well known that cancer cell–endothelial

adhesion is thought to be regulated by the mechanical properties of

the cancer cells and also by the specific expression of various

adhesion molecules and/or ligands to adhesion molecules on the

surface of cancer cells and endothelial cells, and TNF-a-inducible cell

adhesion molecules (CAMs) on the luminal surface of the micro-

vascular endothelium are thought to mediate tumor–endothelial cell

adhesion [29,31,33]. Thus, these results provided preliminary

Fig. 6 – Quantify the effect of TNF-a, anti-VLA-4 as well as anti-VCAM-1 on in vitro fusion. (A) FACS analysis revealed that

bi-fluorescent cells in untreated co-culture of Cal-27 and HUVECs were 0.370.02% (n¼3). (B) TNF-a increased the frequency of

bi-fluorescent cells to 1.6970.09% (n¼3). (C) Anti-VLA-4 and (D) anti-VCAM-1 treatment inhibited the frequency of bi-

fluorescent cells to 0.6070.04% and 0.6470.06%, respectively.

E X P E R I M E N T A L C E L L R E S E A R C H 3 1 8 ( 2 0 1 2 ) 1 7 0 7 – 1 7 1 51714

evidence for the adhesion of VLA-4-positive oral cancer cells to

endothelial cells with enhanced VCAM-1 expression, which is a

prerequisite for cancer–endothelial fusion.

As further evidence for the involvement of VCAM-1/VLA-4

pathway in TNF-a-enhanced cancer–endothelial cell fusion, we

found in current study that both oral cancer–endothelial cell

adhesion and fusions were blocked significantly by anti-VLA-4

antibody or anti-VCAM-1 antibody treatment and reduced to the

control level, suggesting that VCAM-1/VLA-4 was a pair of

adhesion molecule that mediated oral cancer–endothelial cell

adhesion and fusions under TNF-a induced inflammatory condi-

tion. As supportive evidences, Zhang et al. demonstrated that

VCAM-1/a4b1 interaction is the mediators involved in fusion of

Human hematopoietic progenitor cells and murine cardiomyo-

cytes [34]. Okahara et al. also confirmed that interaction between

VLA-4 on tumor cells and VCAM-1 on activated endothelial cells

is critically involved in TNF-a enhancement of metastasis in vivo

[31]. Although the fusion event reduced strikingly, it existed in a

similar level compared to control group after blocking treatment,

which suggested that some other mechanisms may mediate the

fusion of cancer cells and endothelial cells in certain conditions.

Recent report showed that syncytin is expressed by human

cancer cells and is one of the underlying mechanisms of

cancer–endothelial cell fusions and is involved in mediating

fusions between breast cancer cells and endothelial cells [35].

In conclusion, the present study inflammatory cytokine TNF-acould promote oral cancer–endothelial cell fusion. The TNF-a-

enhanced oral cancer–endothelial cell adhesion might be

mediated though VCAM-1/VLA-4 pathway. These findings may

provide insights into tumor–microenvironment interaction and

reveal new therapeutic targets for cancer prevention and

treatment.

Conflict of interest statement

No potential conflict of interest.

Acknowledgments

This work was supported by the National Natural Science

Foundation of China Grant (nos. 30872893 and 81172570) and

the Fundamental Research Funds for Central University Grant

(no. 4103006). Moreover, we thank Shanghai Research Institute

of Stomatology (Affiliated Ninth people’s Hospital, Shanghai

Jiaotong University) for oral squamous cancer cell lines.

r e f e r e n c e s

[1] F. Hillen, A.W. Griffioen, Tumour vascularization: sprouting

angiogenesis and beyond, Cancer Metastasis Rev. 26 (2007)

489–502.

[2] A. Garofalo, R.G. Chirivi, C. Foglieni, R. Pigott, R. Mortarini

I. Martin-Padura, A. Anichini, A.J. Gearing, F. Sanchez-Madrid

E. Dejana, et al., Involvement of the very late antigen 4 integrin

on melanoma in interleukin 1-augmented experimental metas-

tases, Cancer Res. 55 (1995) 414–419.

[3] K. Mortensen, J. Lichtenberg, P.D. Thomsen, L.I. Larsson,

Spontaneous fusion between cancer cells and endothelial cells,

Cell. Mol. Life Sci. 61 (2004) 2125–2131.

E X P E R I M E N T A L C E L L R E S E A R C H 3 1 8 ( 2 0 1 2 ) 1 7 0 7 – 1 7 1 5 1715

[4] I. Singec, E.Y. Snyder, Inflammation as a matchmaker: revisiting

cell fusion, Nat. Cell Biol. 10 (2008) 503–505.

[5] J.M. Nygren, K. Liuba, M. Breitbach, S. Stott, L. Thoren, W. Roell,

C. Geisen, P. Sasse, D. Kirik, A. Bjorklund, C. Nerlov

B.K. Fleischmann, S. Jovinge, S.E. Jacobsen, Myeloid and lym-

phoid contribution to non-haematopoietic lineages through

irradiation-induced heterotypic cell fusion, Nat. Cell Biol. 10

(2008) 584–592.

[6] P.G. Brouckaert, G.G. Leroux-Roels, Y. Guisez, J. Tavernier

W. Fiers, In vivo anti-tumour activity of recombinant human

and murine TNF, alone and in combination with murine IFN-

gamma, on a syngeneic murine melanoma, Int. J. Cancer 38

(1986) 763–769.

[7] N. Watanabe, Y. Niitsu, H. Umeno, H. Sone, H. Neda,

N. Yamauchi, M. Maeda, I. Urushizaki, Synergistic cytotoxic and

antitumor effects of recombinant human tumor necrosis factor

and hyperthermia, Cancer Res. 48 (1988) 650–653.

[8] F. Balkwill, TNF-alpha in promotion and progression of cancer,

Cancer Metastasis Rev. 25 (2006) 409–416.

[9] G. Sethi, B. Sung, B.B. Aggarwal, TNF: a master switch for

inflammation to cancer, Front. Biosci. 13 (2008) 5094–5107.

[10] S.I. Grivennikov, F.R. Greten, M. Karin, Immunity, inflammation,

and cancer, Cell 140 (2010) 883–899.

[11] F. Balkwill, Tumour necrosis factor and cancer, Nat. Rev. Cancer

9 (2009) 361–371.

[12] H. Kobayashi, K.C. Boelte, P.C. Lin, Endothelial cell adhesion

molecules and cancer progression, Curr. Med. Chem. 14 (2007)

377–386.

[13] N. Choe, J. Zhang, A. Iwagaki, S. Tanaka, D.R. Hemenway

E. Kagan, Asbestos exposure upregulates the adhesion of pleural

leukocytes to pleural mesothelial cells via VCAM-1, Am. J.

Physiol. 277 (1999) L292–L300.

[14] J.A. DiVietro, D.C. Brown, L.A. Sklar, R.S. Larson, M.B. Lawrence,

Immobilized stromal cell-derived factor-1alpha triggers rapid

VLA-4 affinity increases to stabilize lymphocyte tethers on

VCAM-1 and subsequently initiate firm adhesion, J. Immunol.

178 (2007) 3903–3911.

[15] S.A. Cannistra, C. Ottensmeier, J. Tidy, B. DeFranzo, Vascular cell

adhesion molecule-1 expressed by peritoneal mesothelium

partly mediates the binding of activated human T lymphocytes,

Exp. Hematol. 22 (1994) 996–1002.

[16] R.R. Langley, R. Carlisle, L. Ma, R.D. Specian, M.E. Gerritsen

D.N. Granger, Endothelial expression of vascular cell adhesion

molecule-1 correlates with metastatic pattern in spontaneous

melanoma, Microcirc. 8 (2001) 335–345.

[17] T. Michigami, N. Shimizu, P.J. Williams, M. Niewolna, S.L. Dallas,

G.R. Mundy, T. Yoneda, Cell–cell contact between marrow

stromal cells and myeloma cells via VCAM-1 and alpha(4)-

beta(1)-integrin enhances production of osteoclast-stimulating

activity, Blood 96 (2000) 1953–1960.

[18] J.K. Slack-Davis, K.A. Atkins, C. Harrer, E.D. Hershey

M. Conaway, Vascular cell adhesion molecule-1 is a regulator

of ovarian cancer peritoneal metastasis, Cancer Res. 69 (2009)

1469–1476.

[19] P. Mattila, M.L. Majuri, R. Renkonen, VLA-4 integrin on sarcoma

cell lines recognizes endothelial VCAM-1. Differential regulation

of the VLA-4 avidity on various sarcoma cell lines, Int. J. Cancer

52 (1992) 918–923.

[20] I. Martin-Padura, R. Mortarini, D. Lauri, S. Bernasconi, F. Sanchez-

Madrid, G. Parmiani, A. Mantovani, A. Anichini, E. Dejana,

Heterogeneity in human melanoma cell adhesion to cytokine

activated endothelial cells correlates with VLA-4 expression,

Cancer Res. 51 (1991) 2239–2241.

[21] J.M. Pawelek, Cancer–cell fusion with migratory bone-marrow-

derived cells as an explanation for metastasis: new therapeutic

paradigms, Future Oncol. 4 (2008) 449–452.

[22] M. Rachkovsky, S. Sodi, A. Chakraborty, Y. Avissar, J. Bolognia,

J.M. McNiff, J. Platt, D. Bermudes, J. Pawelek, Melanoma x

macrophage hybrids with enhanced metastatic potential, Clin.

Exp. Metastasis 16 (1998) 299–312.

[23] X. Lu, Y. Kang, Efficient acquisition of dual metastasis organo-

tropism to bone and lung through stable spontaneous fusion

between MDA-MB-231 variants, Proc. Natl. Acad. Sci. U. S. A.

106 (2009) 9385–9390.

[24] F.R. Miller, A.N. Mohamed, D. McEachern, Production of a more

aggressive tumor cell variant by spontaneous fusion of two

mouse tumor subpopulations, Cancer Res. 49 (1989) 4316–4321.

[25] D.M. Duelli, Y.A. Lazebnik, Primary cells suppress oncogene-

dependent apoptosis, Nat. Cell Biol. 2 (2000) 859–862.

[26] C.B. Johansson, S. Youssef, K. Koleckar, C. Holbrook, R. Doyonnas,

S.Y. Corbel, L. Steinman, F.M. Rossi, H.M. Blau, Extensive fusion

of haematopoietic cells with Purkinje neurons in response to

chronic inflammation, Nat. Cell Biol. 10 (2008) 575–583.

[27] Y. Wu, B.P. Zhou, TNF-alpha/NF-kappaB/Snail pathway in cancer

cell migration and invasion, Br. J. Cancer 102 (2010) 639–644.

[28] R.B. Rebhun, H. Cheng, J.E. Gershenwald, D. Fan, I.J. Fidler,

R.R. Langley, Constitutive expression of the alpha4 integrin

correlates with tumorigenicity and lymph node metastasis of

the B16 murine melanoma, Neoplasia 12 (2010) 173–182.

[29] S. Liang, C. Dong, Integrin VLA-4 enhances sialyl-Lewisx/a-

negative melanoma adhesion to and extravasation through the

endothelium under low flow conditions, Am. J. Physiol. Cell

Physiol. 295 (2008) C701–C707.

[30] M. Klemke, T. Weschenfelder, M.H. Konstandin, Y. Samstag,

High affinity interaction of integrin alpha4beta1 (VLA-4) and

vascular cell adhesion molecule 1 (VCAM-1) enhances migration

of human melanoma cells across activated endothelial cell

layers, J. Cell Physiol. 212 (2007) 368–374.

[31] H. Okahara, H. Yagita, K. Miyake, K. Okumura, Involvement of very

late activation antigen 4 (VLA-4) and vascular cell adhesion

molecule 1 (VCAM-1) in tumor necrosis factor alpha enhancement

of experimental metastasis, Cancer Res. 54 (1994) 3233–3236.

[32] T.C. Wu, The role of vascular cell adhesion molecule-1 in tumor

immune evasion, Cancer Res. 67 (2007) 6003–6006.

[33] H. Li, C. Ge, F. Zhao, M. Yan, C. Hu, D. Jia, H. Tian, M. Zhu, T. Chen,

G. Jiang, H. Xie, Y. Cui, J. Gu, H. Tu, X. He, M. Yao, Y. Liu, J. Li,

HIF-1alpha-activated ANGPTL4 contributes to tumor metastasis

via VCAM-1/integrin beta1 signaling in human hepatocellular

carcinoma, Hepatology (2011).

[34] S. Zhang, E. Shpall, J.T. Willerson, E.T. Yeh, Fusion of human

hematopoietic progenitor cells and murine cardiomyocytes is

mediated by alpha 4 beta 1 integrin/vascular cell adhesion

molecule-1 interaction, Circ. Res. 100 (2007) 693–702.

[35] B. Bjerregaard, S. Holck, I.J. Christensen, L.I. Larsson, Syncytin is

involved in breast cancer–endothelial cell fusions, Cell. Mol. Life

Sci. 63 (2006) 1906–1911.