Alonso et al 23 previously reported that sequence analysis of the human T-cell-specific MAL gene revealed four exons, each encoding a hydrophobic, presumably membrane-associated, segment and its adjacent hydrophilic sequence. Alternative splicing produces four transcript variants, which vary by the presence or absence of alternatively spliced exons 2 and 3. To determine whether alternatively spliced isoforms of MAL were expressed in head and neck squamous epithelium, a pair of PCR-primers spanning exon 1 and exon 4 was designed to discriminate the transcript levels of the four variants using semi-quantitative RT-PCR analysis (Figure 1A, 1B). Our results showed that the variant a, which is the longest of the isoforms is the only transcript variant in head and neck squamous epithelium.

As our microarray data indicated, MAL expression was reduced remarkably in HNSCC as compared to adjacent normal tissues. To investigate whether MAL contributes to HNSCC progression, real-time RT-PCR and semi-quantitative RT-PCR analyses were performed to investigate mRNA levels of MAL in 48 HNSCC specimens. Out of the 48 cases examined, 44 (91.7%) HNSCC specimens had at least a two-fold decrease in MAL mRNA levels as compared to non-tumorous tissues (P < 0.01, Figure 2A, 2B). Meanwhile, the expression levels of MAL mRNA in a panel of nine HNSCC cell lines and the immortalized oral keratinocyte line HIOEC were detected during semi-quantitative RT-PCR analysis. As shown in Figure 2C, MAL transcript levels in the HNSCC cell lines and HIOEC cells were significantly lower than in the case of normal epithelial cells (P < 0.01).

Increasing evidence has shown that epigenetic silencing of tumor suppressor genes because of aberrant DNA hypermethylation and histone modification is essential to carcinogenesis and metastasis 20 24 . To verify whether MAL gene promoter hypermethylation contributed to MAL down-regulation, nine HNSCC cell lines were treated with 5-Aza-dC in conjunction with the TSA. Our result showed that MAL gene transcription was reactivated in 67% (6/9) of the HNSCC cell lines after treatment with 5-Aza-dC, while 89% (8/9) of the cell lines showed MAL reactivation because of simultaneous treatment with 5-Aza-dC and TSA. In addition, the levels of MAL mRNA increased in 44% (4/9) of HNSCC cell lines treated with TSA alone as compared to the untreated control (Figure 3A). To further explore the epigenetic mechanism, the putative CpG islands around MAL transcriptional start site (TSS) were screened by using bioinformatics tools. Two major CpG islands M1 (distance to TSS, -626 to -385 bp) and M2 (distance to TSS, +390 to +669 bp) were found. Next, to evaluate the methylation status of the two major CpG islands around the MAL TSS, bisulfate-treated DNA sequencing was performed in two randomly selected HNSCC cell lines and seven pairs of HNSCCs from 48 cases. The results demonstrated that both CpG islands (that is, M1 and M2) were completely methylated in the HNSCC cell lines (CpG methylated ratio more than 90%) relative to normal primary head and neck epithelial cells (P < 0.01). However, the methylation level of the CpG island (that is, M1) in the MAL putative promoter but not the CpG island in the intragenic regions (that is, M2) was significantly higher in all seven HNSCC cases than in the corresponding adjacent non-cancerous tissues (P < 0.01, Figure 3B, 3C)

To evaluate whether MAL functions as a tumor suppressor in HNSCC cells, we assessed the effect of MAL expression on cell proliferation and colony formation. Based on the MAL expression pattern in HNSCC-derived cell lines (Figure 2C), we transiently transfected mammalian expression vectors containing MAL into Tca, Tca-M and Tb cells, all of which lacked endogenous MAL expression. MAL over-expression suppressed the growth and colony formation ability of these cells relative to mock-transfected controls (P < 0.01, Figure 4A and 4B). To validate which cell population was altered most remarkably in cell cycle distribution, flow cytometry was performed in randomly selected cell line. Our results showed that the MAL-transfected cells had a marked increase in the number of cells in the G1-phase as compared with the mock-transfected cells (43.58% ± 5% vs. 59.61% ± 6%, P < 0.05, Figure 4C.) This indicates that the cells expressing exogenous MAL become blocked at the G1/S transition.

To determine whether the inhibitory effect of MAL expression on cells was caused by apoptosis, two independent assays were performed. As shown in Figure 5A, MAL expression induced an increase in the number of sub-G1 cells. After 24 hours, 48 hours and 72 hours of transient transfection, the apoptotic rates of the MAL-transfected Tca cells were 1.14%, 4.04%, and 11.71%, respectively, as compared with 1.07%, 2.20%, and 4.01%, respectively, for the control cells (48 hours and 72 hours, P < 0.01). In addition, we assessed the fraction of cells with positive staining for 7-amino-actinomycin (7-AAD) and Annexin V-PE in MAL-transfected Tca cells (7.54%) and the control cells (2.60%) (Figure 5B). Furthermore, we also examined the expression level of poly (ADP-ribose) polymerase (PARP) cleavage, which is a hallmark of apoptosis 25 , using Western blot analysis after transient transfection with an empty vector or MAL at 48 hours, 72 hours (Figure 5C).

In vitro invasion assays were performed to determine the effect of MAL on cell invasion using BD BioCoat™Matrigel™Invasion Chambers. The Matrigel™matrix served as a reconstituted basement membrane in vitro. As shown in Figure 6A and 6B, the number of cells that invaded through the transwell membrane in clone #12, which expresses MAL (65 ± 32), was significantly lower than those in clone #6, which does not express MAL (178 ± 44, P < 0.01).

Since MAL expression represses the malignant properties of HNSCC cells in vitro, we also evaluated the effect of MAL on the tumorigenicity in a xenograft model. For this purpose, Tca cells expressing MAL (clone #12) and cells not expressing MAL (clone #6) were injected subcutaneously into the left and right posterior limb of nude mice, respectively. After four weeks, the animals were executed, and the tumor weights were measured. The results showed that tumor growth was significantly reduced in MAL-expressing (clone #12) cells as compared to MAL-non-expressing (clone #6) cells (P < 0.05, Figure 6C).

The samples were collected at the Department of Oral and Maxillofacial Surgery at the Affiliated Ninth People's Hospital of Shanghai Jiao Tong University School of Medicine from 2000 to 2009. All samples were obtained by surgery and then quickly frozen in liquid nitrogen until DNA and total RNA were extracted. Normal tissues adjacent to the tumors were also obtained with the patient's consent. Tumors were classified histologically and staged according to the tumor-node-metastasis (TNM) classification of malignant tumors 39 . Tumor pathological grade was assessed according to standard criteria 40 .

The human cell lines Tca8113 (Tca), Tca8113-M (Tca-M), Tb, Tca/CDDP (established in the Ninth People's Hospital, Shanghai Jiao Tong University School of Medicine), TSCC (kindly provided by Wuhan University, School of Dentistry, China), OSC-4 (kindly provided by Kochi University, School of Medicine, Japan), NB and NT (kindly provided by Nagasaki University School of Dentistry, Japan) were cultured in RPMI-1640 medium (GIBCO BRL, USA) supplemented with 10% heat-inactivated fetal bovine serum (FBS; GIBCO BRL, NY, USA), penicillin (100 units/ml) and streptomycin (100 μg/ml) at 37°C in a humidified 5% CO2 atmosphere. CAL-27 (American Type Culture Collection, Manassas, VA, USA) was cultured in Dulbecco's modified Eagle medium (DMEM; GIBCO BRL, USA) supplemented with 10% heat-inactivated FBS), penicillin (100 units/ml) and streptomycin (100 μg/ml) at 37°C in a humidified 5% CO2 atmosphere. The immortalized oral keratinocyte line HIOEC from primary normal human oral epithelial cells infected with HPV16E6E7 (established in the Ninth People's Hospital, Shanghai Jiao Tong University School of Medicine) was cultured in a defined keratinocyte serum-free medium (KSFM; GIBCO BRL, USA).

Total RNA was extracted from each tissue sample and cell line using the TRIZOL reagent (Invitrogen, San Diego, USA). The quality of the total RNA samples was determined by electrophoresis using formaldehyde agarose gels, and the 18 S and 28 S RNA bands were visualized under ultraviolet light. Then 2 μg of total RNA was reverse-transcribed directly to cDNA using a reverse transcription kit (Promega, Madison, USA) following the manufacturer's instructions in a total volume of 25 μl. The primer sequences used were as follows: MAL forward, 5'-CAGTGGCTTCTCGGTCTT-3'; MAL reverse, 5'-AGCAGAGTGGCTATGTAGGA-3'; β-actin forward, 5'-TCACCCACACTGTGCCCATCTACGA-3'; and β-actin reverse, 5'-CAGCGGAACCGCTCATTGCCAATGG-3'; Each primer was added at a final concentration of 0.5 μM to a 15 μl reaction mixture in PCR buffer containing 1 μl of cDNA, 0.25 mM of each dNTP, 1.5 mM of MgCl₂, and 2.5 units of Taq DNA polymerase. An initial denaturation was carried out for 5 minutes at 94°C, and 35 cycles were performed with the following PCR program: denaturing at 94°C for 30 seconds, annealing at 45-60°C for 30 seconds for MAL and 55°C for 30 seconds for β-actin, and elongation at 72°C for 30 seconds. This program was completed with a final extension at 72°C for 5 minutes. Ethidium bromide-stained bands were visualized using UV transillumination, and fluorescence intensity was quantified using the FR-200 system (FuRi, Shanghai, China). The data from semi-quantitive PCR reactions were normalized against the expression of β-actin from three independent experiments ± the standard deviation (SD). All RT-PCR data were from at least three independent experiments.

All real-time PCR reactions were performed using a Thermal Cycler Dice™Real Time System (Takara) and the SYBR Premix Ex Taq™reagents kit (Takara, Japan). The real-time PCR was performed in a final volume of 15 μl with 1.5 μl of template cDNA at a concentration of 20 ng/μl with 7.5 μl SYBR green I fluorescent dye and 20 pM of each primer for the target gene and the house keeping gene (β-actin). The primer sequences included sense 5'-GGGCTGGGTGATGTTCGT-3' and anti-sense 5'-TAGGCTGCGTCCAAGGTG-3' for MAL (117 bp) and sense 5'-CCTGGCACCCAGCACAAT-3' and anti-sense 5'-GGGCCGGACTCGTCATACT-3' for β-actin. Results of real-time PCR were represented as Ct values, where Ct was a fraction defined as the cycle number at which the sample fluorescent signal passes a given threshold above the baseline. ΔCt was the difference in the Ct values derived from the specific genes after being assayed and β-actin. The N-fold differential expression in a specific gene of a tumor sample, matched normal tissue and cell lines were expressed as 2^ΔCt 17 . The significance level was defined as a P-value of less than 0.01.

To induce demethylation of promoter prior to evaluation for induction of MAL expression, all of the above-mentioned HNSCC cell lines were treated with 2 μM 5-Aza-dC, which is the DNA demethylation reagent, for 72 hours, 0.5 μM TSA, which is a specific inhibitor of histone deacetylase, for 24 hours, or 2 μM 5-Aza-dC for 72 hours plus 0.5 μM TSA for 24 hours.

Genomic DNA was treated with bisulfate as previously described 18 . Briefly, 1 μg genomic DNA was denatured by incubation with 0.2 M NaOH. Aliquots of 10 mM hydroquinone and 3 M sodium bisulfate (pH = 5.0) were added, and the solution was incubated at 50°C for 16 hours. To analyze the DNA methylation status of the CpG islands of MAL promoter in HNSCCs and cell lines, regions enriched in CpG islands were amplified in bisulfate-treated genomic DNA using the primers CpG island I, forward, 5'-GGAGTAATTTTTTATTTTTAGGTAGA-3'; reverse, 5'-AAATTTAAATCTCCTTCATTTTTCC-3'; CpG island II, forward, 5'-TTTAATTGGGGTTAGATGTAGGTAG-3'; and reverse, 5'-AAAAACTTTAAAAAACCAAAAAAAA-3'. The PCR products were then subcloned into a pMD18-T vector (TaKaRa Inc. Japan) for DNA sequencing on an ABI 3730 sequencer.

The cells were washed with PBS twice and lysed for 30 minutes in 25 mM Tris-HCl with a pH of 7.5 and containing 150 mM NaCl, 5 mM EDTA, and 1% Triton X-100 at 4°C. The lysate was homogenized by passing the sample through a 22-gauge needle. The glycolipid-enriched membrane (GEM) was isolated from the lysate by equilibrium centrifugation for Western blot analysis with an anti-c-myc monoclonal antibody (mAb). In brief, for immunoblot analysis, the GEM samples were run on SDS-PAGE in 12% acrylamide gels and transferred onto nitrocellulose membranes (Amersham, NA, England). After blocking with 5% nonfat dry milk and 0.05% Tween-20 in PBS, blots were incubated with anti-c-myc mAb (Santa Cruz Biotechnology, CA, USA), anti-cleavage PARP polyclonal antibody (Cell signaling, MA, USA) and anti-β-actin mAb (Sigma, St Louis, MO, USA) at the optimized dilutions. Bands were detected using an IRDye™800 Conjugated Affinity Purified Anti-Mouse IgM antibody (Rockland, Gilbertsville, PA, USA). The membrane was then washed several times and scanned using the Odyssey infrared imaging system (LI-COR, Lincoln, NE, USA) at an 800 channel wavelength and analyzed with Odyssey software.

To generate the MAL expression vector, the open reading frame of human MAL cDNA was cloned into the eukaryotic expression vector pcDNA3.1 (Invitrogen, San Diego, CA, USA) and fused to a COOH-terminal Myc tag. The primers used for the amplification of the open-reading frame of the MAL cDNA were GCGAATTCACCATGGCCCCCGCAGCGGCGACG (forward primer), with an EcoRI site, and GCAAGCTTGCTGAAGACTTCCATCTGATTAAAG (reverse primer), with a HindIII site without the stop codon. The amplified MAL gene product was purified using a Qiaquick PCR Purification Kit (Qiagen, Chatsworth, CA, USA), cut with EcoRI and HindIII and ligated with the Ligation Mix (TakaRa, Dalian, China) into the respective EcoRI and HindIII sites on the eukaryotic expression vector pcDNA3.1. For transient transfection, the cells were transfected with various plasmids at about 80% confluence using the Lipofectamine™2000 (Invitrogen, San Diego, CA, USA) reagent according to the manufacturer's instructions. To generate stable MAL-expressing cell lines, cells were selected with G418 (600 μg/ml) for two weeks, and drug-resistant cells were screened using Western blot analysis with the anti-human c-myc mAb to detect MAL-myc. The clones that exhibited positive immunoreactivity were maintained in G418 (300 μg/ml) medium. Mock-transfected, MAL-transfected (but not MAL-expressing) and MAL-expressing stable cells were denoted as 3.1-vector, clone #6, and clone #12, respectively.

Cell-proliferation assay was performed to analyze the proliferation potential of transiently transfected empty vector and MAL-transfected cells. To this end, we used the Cell-Counting Kit (CCK)-8 (Dojindo, Kumamoto, Japan). Briefly, the cells were harvested and plated in 96-well plates at 1 × 10³ cells per well and maintained at 37°C in a humidified incubator. At the indicated time points, 10 μl of the CCK-8 solution was added into the triplicate wells and incubated for 1 hour, and the absorbance at 450 nm was measured to calculate the number of vital cells in each well. Measurements were performed in triplicate, and the Mean (±SD) optical density (OD) was reported.

Twenty-four hours after transfection, the cells (1 × 10⁵ cells per plate) were plated in 100-mm culture dishes and incubated with 600 μg/ml of G418 for 14 days to allow for colony formation. The colonies were then washed twice with PBS, fixed with 70% ethanol and stained with 0.1% Coomassie Brilliant Blue R-250. Colonies of more than 50 cells were counted under a dissecting microscope. The data from colony formation were shown as means (± SD) from at least three independent experiments, each performed in triplicate.

Empty vector and MAL-transfected cells at the logarithmic growth phase were harvested, and single-cell suspensions containing 1 × 10⁶ cells were permeabilized with 70% ethanol. The cells were then labeled with 50 μg/ml of propidium iodide (PI) and treated with 250 μg/ml of RNase at 4°C for 30 minutes. Analysis was performed using FACS Calibur (BD Biosciences, San Jose, CA, USA) and analyzed using Cell Quest Software (BD Biosciences, San Jose, CA, USA). The data represent the Means (± SD) from at least three independent experiments.

The cells were harvested 24 hours, 48 hours, and 72 hours after transient transfection. Then they were washed once in PBS, permeabilized with 70% ethanol for 20 minutes and stored at -20°C overnight. The cells were then subsequently washed with PBS and stained with 50 μg/ml of propidium iodide at 4°C for 30 minutes. Apoptotic cells were assessed by flow cytometry for sub-G1 DNA content. Meanwhile, apoptosis was assessed using Annexin V-PE staining 72 hours after transfection. Then, these cells were quantified by flow cytometry using the Annexin V-PE Apoptosis Detection Kit (BD Biosciences, San Jose, CA, USA) according to the manufacture's protocols. Briefly, floating cells and trypsinized adherent cells were pooled and resuspended in 100 μl of Annexin V-PE Binding Buffer. Then 5 μl of Annexin V-PE and 5 μl of 7-AAD were added, and the cells were incubated for 15 minutes at 25°C. The cells were then resuspended in 400 μl of the Annexin V-PE binding buffer and analyzed immediately by flow cytometry. Then 10,000 events were scored by the FACSCalibur and analyzed using the Cell Quest software.

Stable cells, that is, clone #6 and clone #12, were grown on coverslips and washed once with ice-cold PBS and fixed for 30 minutes in 4% paraformaldehyde at 4°C. For immunostaining, the fixed cells were permeabilized with 0.1% Triton X-100 in PBS for 5 minutes, washed and blocked with PBS containing 0.05% Tween-20 and 5% horse serum (HS) for 30 minutes at room temperature, and incubated with a 1:500 dilution of anti-c-myc mAb in PBS supplemented with 0.05% Tween-20 and 5% HS in a humidified chamber at 4°C overnight. The cells were then rinsed three times with 0.05% Tween-20 in PBS and incubated with a 1:100 dilution of secondary antibody for 30 minutes in 0.05% Tween-20 and 5% HS in PBS. The cells were then rinsed three times with 0.05% Tween-20 in PBS. All fluid was then removed, and two separate drops of Gel Mount-4' and 6' diamino-2-phenylindole·2HCl (DAPI) were applied to the sections. A cover slip was then carefully lowered onto the sections. The cells were then imaged using a TCS-SP laser-scanning confocal microscope with a ×40 oil immersion lens (Leica Microsystems, Mannheim, Germany).

In vitro invasion assay was performed to analyze the invasive potential of stable clone #6 and clone #12 cells. A total of 1 × 10⁵ cells in 750 μl of serum-free RPMI-1640 medium were plated onto BD BioCoat ™Matrigel ™Invasion Chambers (8 μm pore size; BD Biosciences, USA) with the lower chamber containing 750 μl of RPMI-1640 medium with 10% FBS as a chemoattractant. After 48 hours of incubation in a humidified atmosphere containing 5% CO₂ at 37°C, the non-invading cells were removed from the upper surface of the membrane by a cotton swab. The membranes were then fixed with methanol and stained with 0.5% crystal violet stain. Invading cells were photographed and counted in five random non-overlapping 200× fields under a light microscope.

Nude mice (4-6 weeks old) were used for xenograft studies. Tca cells (1.5 × 10⁶ in PBS) expressing MAL (clone #12) or not (clone #6) were injected subcutaneously in the extremities of the nude mice. After four weeks, nude mice bearing tumors were executed, and tumor weight was measured using an electronic balance. The values for nine mice in each group were averaged to obtain the mean ± (SD) Differences in tumor weights were analyzed using the Student's t-test.

Semi-quantitative RT-PCR, real-time PCR and in vitro statistical analyses were performed with software from SPSS 13.0 for Windows (Chicago, IL, USA). Results of the semi-quantitative RT-PCR and real-time PCR analyses were evaluated using the Mann-Whitney test for two independent groups. The results of the cell proliferation assay, colony formation assay, and in vitro invasion assay were evaluated using Student's t-tests. A P-value of less than 0.05 was considered significant.