Arsenic trioxide reduces the invasive and metastatic properties of nasopharyngeal carcinoma cells in vitro

Nasopharyngeal carcinoma (NPC) is notorious for the metastases, which are in close association with Epstein-Barr virus-encoded latent membrane protein 1 (LMP1). Arsenic trioxide (As2O3) has been shown to induce apoptosis and differentiation in NPC xenografts. Then, can it repress the cancer cells’ metastasis potential? To elucidate this issue, the present study was performed. LMP1-negative cell line HNE1 and LMP1-positive cell line HNE1-LMP1 were used as in vitro model. Cells (1 x 105/mL) were cultured with or without 3 μM As2O3 for 48 h. Then the survival cells were collected to investigate their potential of colony formation, attachment, invasion, and migration. Both confocal immunofluorescence staining and Western blot were used to detect the changes of LMP1 expression. The changes of MMP9 were examined by RT-PCR assay and Western blot. The results were as follow: i) the colony formation inhibition rate (75.41 ± 3.9% in HNE1-LMP1 cells vs 37.89 ± 4.9% in HNE1 cells), the rate of attachment (HNE1-LMP1 vs HNE1: 56.40 ± 3.5 vs 65.87 ± 5.9%), the invasion inhibitory rate (HNE1-LMP1 vs HNE1: 56.50 ± 3.7 and 27.91 ± 2.1%), and the migration inhibitory rate (HNE1-LMP1 vs HNE1: 48.70 ± 3.9 vs 29.19 ± 6.27%) were all significantly different between the two cell lines (P < 0.01). ii) LMP1 was down-regulated in As2O3-treated HNE1-LMP1 cells. iii) The reduction of MMP-9 was found in As2O3-treated groups, more evident in HNE1-LMP1 cells. Thus, we conclude that As2O3 can reduce metastasis potential of NPC cells, involving inhibition of MMP-9 expression. LMP1 were also reduced in this process and seemed to enhance anti-metastasis activity of As2O3. Correspondence


Introduction
Nasopharyngeal carcinoma (NPC) is a common cancer in southern China.The disease is notorious for its high invasiveness and metastatic activity.Clinically, tumor cells often disseminate to regional lymph nodes and to distant sites before forming a mass at the primary site.Thus far, radiotherapy or radiochemotherapy is the treatment of choice for this malignancy, but side effects as gastrointestinal toxicity and myelosuppression often lead to the interruption of chemotherapy.In addition, tumor metastasis remains a critical obstacle in clinical radiotherapy or combined radiochemotherapy.When metastatic disease develops after curative radiotherapy, the prognosis is poor.Therefore, a novel therapeutic approach to NPC is strongly desired.
We have reported that arsenic trioxide (As 2 O 3 ), at the dose of 5 mg/kg, suppressed the growth of NPC xenografts by inducing partial differentiation and apoptosis (1).In our earlier work, the drug was shown to inhibit telomerase activity and to enhance radiation-induced apoptosis (2,3).However, the effects of As 2 O 3 on NPC cell invasion and metastasis are unclear.In the present study, we investigated the antimetastatic effects of As 2 O 3 on human NPC cell lines using the Matrigel invasion assay.In view of the important role of latent membrane protein 1 (LMP1) in NPC metastasis, we determined the influence of LMP1 on As 2 O 3mediated anti-metastatic activity.

Cell lines
The HNE1-LMP1 cell line, which is constantly expressing LMP1 by artificially transfecting LMP1 cDNA into HNE1 cells, was investigated as an in vitro model in our study.The parental cell line HNE1, which does not express LMP1, derived from poorly differentiated NPC, was used as control.The two cell lines were established at the Cancer Research Institute of Hunan Medical University (4) and were kindly provided by Professor Y. Cao.Cells were cultured in RPMI 1640 medium with 10% fetal bovine serum, kept in a humidified atmosphere of 95% air and 5% CO 2 at 37ºC.

Experimental design
Both cell lines, at the density of 1 x 10 5 / mL, were exposed to 3 µM As 2 O 3 for 48 h.Floating cells were then discarded and the residual cells were allowed to grow further.When the residual cells reached confluence they were collected for study.Cells not treated with As 2 O 3 were used as controls.

Colony formation assay
About 100 living cells were added to a 60-mm culture dish containing 5 mL of culture medium.The dishes were then placed in a humidified incubator containing 5% CO 2 and incubated at 37ºC for 14 days.The number of colonies containing more than 50 cells in each dish was counted under a microscope.The inhibitory rate (IR) was calculated as follows: IR (%) = (number of colonies formed in the control group -number of colonies formed in the test group)/number of colonies formed in the control group.

In vitro adhesion assay
Each well of the 96-well microplates was coated with reconstituted basement membrane Matrigel (Becton Dickinson Labware, Franklin Lakes, NJ, USA), 2 µg per well.The coated wells were allowed to dry and appropriate serum-free RPMI 1640 medium was added to each well and incubated for 1 h.The wells were then washed with PBS.Each well containing 4 x 10 4 cells was incubated for 2 h at 37ºC in the presence of 5% CO 2 .Wells were washed 3 times with 200 µL PBS, 40 µg MTT was added to each well and the plates were incubated for 4 h at 37ºC in the presence of 5% CO 2 .The liquid was then removed and 200 µL DMSO was added to each well.The absorbance of each well was read with a microplate reader (Bio-Rad 450, Hercules, CA, USA) at 492 nm.Data are reported as the percentage of total cells, assuming that the adhesion of cells in the control treatment represented 100%.

In vitro invasion assay
Assays were performed with the use of Falcon cell culture inserts (pore size, 8.0 µm; Becton Dickinson).The chambers were set in a 6-well plate.The upper layer of the culture insert was then coated with 750 µg Matrigel, a reconstituted extracellular matrix (Becton Dickinson).Cells were seeded at a density of 2 x 10 4 cells/well into the upper layer of the culture insert and cultured with serum-free DMEM.Then, 3 mL of culture medium supplemented with 0.1% BSA and 250 µg solubilized Matrigel was placed into the lower layer of the culture insert as a chemoattractant.After the cells were incubated for 24 h, the remaining cells in the upper layer were swabbed with cotton and penetrating cells in the lower layer were fixed with 95% ethanol and removed for hematoxylin staining.Cells passing through the Matrigel matrix and each 8-µm pore of the culture insert were counted using light microscopy.Ten fields per well were counted.The IR was calculated as follows: IR (%) = (number of penetrating cells in the negative control group -number of penetrating cells in the test group)/number of cells in the penetrating negative control group.

In vitro mobility assay
The assay was the same as for the invasive procedure described above except that the upper side of the polycarbonate membranes was not coated with Matrigel.

LMP1 expression assay
To determine whether LMP1 plays a role in the metastatic potential of the cells we first used confocal immunofluorescence staining to detect the changes in LMP1 expression in HNE1-LMP1 cells after As 2 O 3 treatment.Cells were cultured overnight on a glass coverslip, washed with PBS and fixed in 4% paraformaldehyde.To display LMP1, the cells were first incubated with anti-LMP1 antibody (DAKO, Glostrup, Denmark) and then reacted with their corresponding FITC-conjugated anti-IgG antibody (DAKO) as secondary antibody.To visualize the nuclei, the cells were stained with 50 µg/mL propidium iodine containing 100 µg DNase-free RNase A per mL and the fluorescent image was observed under a laser-scanning confocal microscope (Ultima 312, Meridian Instruments Inc., Kent, WA, USA) using the following parameters: excited light 488 nm, emission light 530 nm and pinhole 10-40 nm.Western blotting was then used to analyze the changes in LMP1 expression.The procedure was similar to that used for matrix metalloproteinase 9 (MMP-9) examination, which is described below.
In addition, SDS-PAGE and Western blot were used to determine whether MMP-9 expression was affected.Cells were collected with a cell scraper and placed in lysis buffer (1 mM Na 3 PO 4 , 20 mM HEPES, pH 7.4, 1 mM PMSF, 10 µg/mL leupeptin, and 10 µg/ mL aprotinin) for at least 1 h at 4ºC.The cells were transferred to a clean tube and centrifuged for 30 s at 1,500 g to eliminate unbroken cells.The supernatant was incubated on ice for 1 h and then centrifuged at 10,000 g at 4ºC for 5 min and the supernatant was collected and placed in a new microcentrifuge tube.The protein was stored at -70ºC.Equal amounts of protein from each cell line were added to 12.5% polyacrylamide gel.After electrophoresis, proteins were blotted onto PVDF sheets (Amersham Pharmacia Biotech Inc., Piscataway, NJ, USA) at 350 mA for 1 h.The same antibodies used for immunofluorescence staining were used to detect MMP-9 in each tumor cell line.Goat antimouse antibody conjugated with horseradish peroxidase was used as the secondary antibody.The protein bands were visualized with ECL plus Western blotting detection reagents (Amersham).

Statistical analysis
Each assay was performed in triplicate and repeated at least three times.Statistical analysis was performed using the SPSS 10.0 for Windows package.Data are reported as means ± SD.Statistical differences were evaluated by a one-way analysis of variance (ANOVA) followed by the Bonferroni post hoc test, with the level of significance set at P < 0.05.

Effect of As 2 O 3 on colony formation
After As 2 O 3 treatment, the two cell lines showed different rates of colony formation (see Figure 1).The rate of colony formation inhibition was 75.41 ± 3.9% for HNE1-LMP1 cells and 37.89 ± 4.9% for HNE1 cells, with a significant difference between lines (P = 0.000).

Effect of As 2 O 3 on the invasive ability of the cells
As shown in Figure 3, there were fewer invading HNE1-LMP1 and HNE1 cells in the As 2 O 3 treatment group compared to the

Effect of As 2 O 3 on adhesion
The influence of As 2 O 3 treatment on the adhesion of both cell lines is shown in Figure 2.Both lines showed significantly decreased attachment to Matrigel, with values of 56.40 ± 3.5% for HNE1-LMP1 cells and 65.87 ± 5.9% for HNE1 cells (P = 0.001).

Effect of As 2 O 3 on cell chemotactic migration ability
The chemotactic ability of both cell lines after treatment with As 2 O 3 was lower than

E
that of their negative control group (HNE1-LMP1: P = 0.000; HNE1: P = 0.000), but the inhibitory rate of HNE1-LMP1 (48.70 ± 3.9%) was significantly higher than that of the HNE1 group (29.19 ± 6.27%; P = 0.000).The effects of As 2 O 3 on the chemotactic migration ability of both cell lines are shown in Figure 4.

Effect of As 2 O 3 on LMP1 expression
As shown in Figure 5, the positive staining intensity of HNE1-LMP1 cells markedly decreased after As 2 O 3 treatment, as demonstrated by confocal microscopy.In parallel, Western blot showed that the level of LMP1 was also reduced.The level of MMP-9 mRNA was examined by semi-quantitative RT-PCR in the HNE1-LMP1 and HNE1 cell lines.As shown in Figure 6, MMP-9 mRNA was expressed in both HNE1-LMP1 and HNE1 cells.However, the extent of expression became very faint in HNE1-LMP1 cells after As 2 O 3 treatment, whereas As 2 O 3 -treated HNE1 cells showed only moderate reduction of MMP-9 mRNA expression.In addition, tumor cells showed remarkable alteration in the expression of MMP-9 protein.In both cell lines, the expression of MMP-9 protein (92 kDa) decreased after treatment with As 2 O 3 .The down-regulated protein expression of MMP-9 was observed particularly in As 2 O 3 -treated HNE1-LMP1 cells (Figure 7).

Discussion
NPC has highly invasive and metastatic properties and is more metastatic than other head and neck carcinomas.Approximately 90% of patients show cervical lymph node metastases as the most frequent finding of NPC.At present, radiotherapy remains a standard treatment for this disease.Although intensity modulation radiotherapy delivers a higher conformal radiation dose to the target area and may spare normal organs such as the parotid glands in patients with early stage disease, distant metastases still represent the predominant mode of treatment failure (6,7).Thus, there is a great need for effective antimetastasis treatment for patients with NPC.
As 2 O 3 , a component of Chinese medicine, has been known as a poison for many years.The drug has started drawing more attention since the discovery of its clinical efficacy in acute promyelocytic leukemia reported by Chinese investigators.Clinical trials administering As 2 O 3 against other blood malignancies and even solid tumors have been conducted in recent years (8).FDA approval of As 2 O 3 against acute promyelo-  cytic leukemia without major clinical complications has quickly led to the testing of As 2 O 3 against several types of cancers (9), even though the complete mechanism of action against solid or blood-borne tumors is unclear.At present, the drug is indicated as a broad-spectrum anticancer medicine for a variety of cancers.We have reported that single doses of 5 mg/kg As 2 O 3 cause apoptosis and differentiation in NPC xenografts (1,2).Moreover, recent data indicated that As 2 O 3 can induce vascular shutdown and necrosis in esophageal carcinoma (10).In addition, As 2 O 3 has also been reported to inhibit radiation-induced cell invasion (11).Considering these data as a whole, it was reasonable to investigate whether As 2 O 3 is able to inhibit the invasive and metastatic activity of NPC cells.
Because we know that not all NPC cells can be induced to apoptosis or differentiation after As 2 O 3 treatment, the objective of the present study was to determine whether the malignant biological behavior of residual cells can be changed.Therefore, the present experiment was designed to further extend our previous studies of the inhibitory effects of As 2 O 3 on cancer cells in NPC, with emphasis on the following points: i) the potential metastatic activity of residual NPC cells after As 2 O 3 treatment and its possible mechanism, and ii) the role of LMP1 in the anticancer effect of As 2 O 3 .
It is well known that tumor cell metastasis is a complex cascade of events.Essential steps include the degradation of extracellular matrix and basement membrane.The process involves multiple steps such as proliferation, adhesion and migration of tumor cells (11).In the present study, we found that As 2 O 3 inhibited the cell potential for proliferation, attachment, invasion, and migration.Previous studies have emphasized that the critical step for the control of metastases is to retard local proliferation (12) and to block cell attachment, invasion and motility.Current data indicate that 3 µM As 2 O 3 for 48 h can suppress clonogenic survival, especially in HNE1-LMP1 cells.It has been shown that cellular interactions with extracellular matrix will promote adhesion and migration, which are thought to be required for tumor metastases.Agents inhibiting cell attachment in vitro may decrease the invasion and metastatic potential of tumor cells in vivo (13).In vitro invasion, attachment, proteolytic dissolution of the matrix, and movement of tumor cells through Matrigel and polycarbonate are required.Thus, the reconstituted basement membrane invasion assay could reflect the invasive ability of the tumor.We reported here that, compared with untreated cells, the number of residual HNE1-LMP1 and HNE1 cells after the action of As 2 O 3 treatment through the membrane decreased by up to 1-to 2-fold.This is the first evidence that As 2 O 3 can reduce the invasive potential of NPC tumor cells in vitro.
NPC cells have a unique environment, because most carcinoma cells contain hu-man Epstein-Barr virus as a major etiologic agent for carcinogenesis.Epstein-Barr virus-encoded viral oncoprotein, LMP1, cooperatively induces cellular immortalization and transformation by a series of signal transductions (12).The LMP1 up-regulates vascular endothelial growth factor, cyclooxygenase-2, and interleukin-8 (13-15) that is actively involved in the promotion of angiogenesis.We know that massive formation of blood vessels at the tumor site will increase the opportunity for tumor cells to enter the circulation.Our data showed that As 2 O 3 inhibited LMP1 expression in HNE1-LMP1 cells.As 2 O 3 -induced metastasis suppression occurred more easily in LMP1-positive NPC cells (HNE1-LMP1) than in the parental cells (HNE1) with no LMP1 expression.Whether microvessel synthesis was inhibited accordingly remains to be determined.
In the present study, we examined the alteration of MMP-9 mRNA levels by RT-PCR methods and its protein expression by Western blot, because MMP-9 belongs to a gene family of zinc-containing endopeptidases which can degrade the extracellular matrix and basement membrane, playing an essential role in the metastatic process.On the other hand, many studies have been published about the association of LMP1 and MMP-9 with the invasive and metastatic potential of NPC (13,16).Our data indicated that As 2 O 3 can down-regulate MMP-9 at the mRNA and protein level in both cell lines to some extent, but with strong inhibition in the HNE1-LMP1 cell line.Based on these findings, we suggest that LMP1 participates in the enhancement of MMP-9 suppression induced by As

Figure 2 .Figure 3 .
Figure 2. Effect of As 2 O 3 (3 µM) for 48 h on adhesion in both cell lines.Absorbance of each well was obtained at 492 nm.A, Absorbance for HNE1-LMP1 cells without As 2 O 3 treatment.B, Absorbance for HNE1-LMP1 cells after As 2 O 3 treatment.C, Absorbance for HNE1 cells without As 2 O 3 treatment.D, Absorbance for HNE1 cells after As 2 O 3 treatment.Absorbance at 492 nm 2.5 2.0 1.5 1.0 0.5 0 A B C D

Figure 5 .
Figure 5.Effect of 3 µM As 2 O 3 for 48 h on LMP1 expressed in HNE1-LMP1 cells.Top panel, Demonstration of LMP1 expression by confocal immunofluorescence staining.The positive signals were green-stained and located in the cytoplasm and membrane (400X).A, Strong green fluorescent intensity for LMP1 signals in untreated cells.B, Reduction of green fluorescent intensity for LMP1 signals in As 2 O 3 -treated cells.Bottom panel, Western blott analysis of LMP1 proteins.A, Strong positive staining of LMP1 expression in the control group.B, Weak LMP1 expression in the As 2 O 3 group.

Figure 4 .
Figure 4. Effect of 3 µM As 2 O 3 for 48 h on the migration potential of cells.The small rings are 8-µm membrane pores of the Falcon cell culture inserts indicated by a solid arrow.The migration cells indicated by an open arrow were counted at least in 10 fields per insert.A, Penetrating HNE1-LMP1 cells in the negative control.B, Penetrating HNE1-LMP1 cells in the As 2 O 3 group.C, Penetrating HNE1 cells in the negative control.D, Penetrating HNE1 cells in the As 2 O 3 group.E, Number of penetrating cells per insert (y-axis) in each group.Data are reported as the mean ± SD for three independent experiments carried out in triplicate, where A, B, C, and D in the x-axis stand for each experimental group described as above.H&E staining.P < 0.05 compared to control (ANOVA).

Figure 6 .
Figure 6.Effect of 3 µM As 2 O 3 for 48 h on MMP-9 mRNA expression in HNE1-LMP1 and HNE1 cells.M, DNA markers.A, Untreated HNE1-LMP1 cells.B, As 2 O 3 -treated HNE1-LMP1 cells.C, Untreated HNE1 cells.D, As 2 O 3 -treated HNE1 cells.The vertical scale at the left is bp.Experiments were performed on three different occasions with similar results.

2 O 3 .
Further studies are needed to clarify how As 2 O 3 regulates the expression of MMP-9 and how LMP1 increases the sensitivity of As 2 O 3 in altering the invasive and metastatic properties of NPC cells.We analyzed the antiproliferative and antimetastatic effects of As 2 O 3 in NPC cell lines.The down-regulation of MMP-9 by As 2 O 3 may result in lower invasiveness by NPC cells.Our data also demonstrate that LMP1 expression enhances the responsiveness of tumor cells to As 2 O 3 .Based on these results, we present a new point of view about