The study protocol was approved by the Ethics Committee of the Affiliated Hospital of Nantong University (Nantong, China). A total of 20 pairs of OSCC tissue and adjacent healthy tissue samples were collected between March 2015 and April 2017 at the Affiliated Hospital of Nantong University. All the patients provided written informed consent for their tissues to be used for research purposes. None of the patients had received radiation therapy prior to surgical resection. The collected tissues were immediately snap-frozen in liquid nitrogen for later use.

The OSCC cell lines HSC-3, Ca9-22, SAS and OSC-19 were purchased from American Type Culture Collection. The HSC-3, SAS and OSC-19 cell lines were cultured in Dulbecco's modified Eagle's medium (DMEM; Sigma-Aldrich; Merck KGaA) supplemented with 10% fetal bovine serum (FBS; HyClone; GE Healthcare Life Sciences), and Ca9-22 cells were maintained in DMEM/F-12 (HyClone; GE Healthcare Life Sciences) with 10% FBS at 37°C with 5% CO₂ in a CO₂ incubator. The 293 cell line, a vehicle for the production of adenoviral vaccines and recombinant proteins (19), was purchased from the Type Culture Collection of the Chinese Academy of Science and maintained in modified Eagle's medium (MEM; HyClone; GE Healthcare Life Sciences) supplemented with 10% FBS (HyClone; GE Healthcare Life Sciences).

The interfering RNA for MSI1 (sh-MSI1) and negative plasmid (sh-Ctrl) were purchased from General Biosystems Co., Ltd. To construct the vector containing coding sequences of MSI1, cDNA was reverse-transcribed from total RNA extracted form HSC-3 cells, amplified with PCR with MSI1 primers and then cloned into the pcDNA3.1 vector. Approximately 1×10⁶ cells/per well were seeded and grown overnight in 6-well plates. On the following day, Lipofectamine 2000 reagent (Thermo Fisher Scientific, Inc.). was used for transient transfection of the cells with 3.0 μg plasmids, including sh-Ctrl, sh-MSI1, MSI1 overexpression plasmids (MSI1) or negative empty pcDNA3.1 vector (pcDNA3.1). At 48 h post-transfection, the cells were harvested for western blotting, reverse transcription-quantitative pcr (RT-qPCR) analysis, flow cytometry or other experiments at the indicated times.

For in vivo experiments, a lentiviral vector carrying interfering RNA for MSI1, pLKD-U6-MSI1-shRNA or a negative control vector, pLKD-U6-shRNA, and corresponding viruses (1×10⁸ pfu) were custom-constructed and prepared by OBiO Technology (Shanghai) Corp., Ltd.

The plasmids containing firefly luciferase reporters and MSI1-silencing plasmids were co-transfected into 293 cells using Lipofectamine 2000 (Thermo Fisher Scientific, Inc.). The cells were then lysed at 48 h after transfection and examined using a dual-luciferase assay (Promega Corporation) according to the manufacturer's instructions. Luciferase activity was expressed as the ratio of firefly to Renilla luciferase activity.

Approximately 5,000 cells were plated in 96-well plates with 200 μl medium per well. The cells were then transfected as indicated with sh-MSI1 (MSI1 suppression plasmid), negative control vector (pcDNA 3.1 or sh-Ctrl) or MSI1 overexpression plasmid (MSI1) for 48 h, after which time 2 μl CCK-8 solution (Biosharp) was added to each well and the cells were incubated for a further 4 h. The absorbance at 450 nm was measured using a micro-plate reader (Thermo Fisher Scientific, Inc.).

To detect BrdU incorporation, the cells transfected with the indicated plasmids were washed thoroughly with medium and then cultured in fresh medium containing 10 μM BrdU (Sigma-Aldrich; Merck KGaA) for 1 h. The cells were then allowed to grow in BrdU-free medium for another 48 h, after which time they were harvested for detection. The harvested cells were washed with PBS, fixed in 70% ice-cold ethanol, resuspended in 2N HCl and incubated for 30 min at room temperature, followed by hybridization with a mouse monoclonal anti-BrdU antibody (cat. no. ab8152, dilution, 1:500, Abcam) overnight at 4°C. Finally, the cells were rinsed with PBS combined with Tween-20 and incubated with fluorescein isothiocyanate-conjugated rabbit anti-mouse immunoglobulin antibody (Jackson ImmunoResearch), followed by staining with 4′,6-diamidino-2-phenylindole solution for 10 min prior to image capture.

RNA was isolated from tissues and cell lines using TRIzol reagent (Invitrogen; Thermo Fisher Scientific, Inc.). First-strand cDNA was synthesized using a TIANScript RT kit (Tiangen Biotech). Subsequently, the MSI1 expression levels were measured using SYBR-Green™ PCR Master Mix (Invitrogen; Thermo Fisher Scientific, Inc.), with β-actin serving as an endogenous control. The data were analyzed using the 2^−ΔΔCq method (20). The primer sequences for the genes were as follows: MSI1: Forward, 5′-GATCCAGGGGTTTCGGCTTC-3′ and reverse, 5′-GAAGGCCACCTTAGGGTCAA-3′; β-actin: Forward, 5′-GATGAGATTGGCATGGCTTT-3′ and reverse, 5′-GTCACCTTCACCGTTCCAGT-3′.

Transwell 24-well filters (pore size, 8 μm; BD Biosciences) were precoated with Matrigel at 37°C for 30 min. The cells transfected with the indicated plasmids were starved in serum-free medium overnight, and then suspended in medium containing 2% FBS. Approximately 20,000 cells were added to the upper chamber of the filters, and medium containing 20% FBS was added to the lower chamber. After incubating for 24 h, the cells on the lower surface of the membrane were fixed with cold methanol and stained with 0.1% crystal violet solution. For the migration assay, the cells were plated on Transwell 24-well filters in plates without Matrigel, and the protocol was the same as that used in the invasion assays described above. Finally, the cells in at least five random microscopic fields were counted and photographed.

Cell lysates from patient tissues and transfected cells were prepared with radioimmunoprecipitation assay lysis buffer (Beyotime Institute of Biotechnology), and protein concentrations were quantified using a bicinchoninic acid assay kit (Biosharp). Using electrophoresis on 10% sodium dodecyl sulfate-polyacrylamide gel, a total of 30 μg protein was separated and transferred to a polyvinylidene difluoride membrane (EMD Millipore). The membrane was then probed with antibodies against rabbit monoclonal pS727-STAT3 (cat. no. ab32143, dilution, 1:8,000), rabbit monoclonal STAT3 (cat. no. ab32500, dilution, 1:1,000), rabbit monoclonal c-Myc (cat. no. ab32072, dilution, 1:1,000), rabbit monoclonal cyclin D1 (cat. no. ab134175, dilution, 1:25,000), rabbit monoclonal MSI1 (cat. no. ab52865, dilution, 1:1,000), rabbit monoclonal p21 (cat. no. ab109520, dilution, 1:5,000), rabbit monoclonal p27 (cat. no. ab32034, dilution, 1:5,000) or mouse monoclonal β-actin (cat. no. ab6276, dilution, 1:6,000) at 4°C overnight. All antibodies were purchased from Abcam Biotechnology. After washing with a mixture of Tris-buffered saline and Tween 20, the membranes were incubated with horseradish peroxidase-conjugated anti-mouse or anti-rabbit antibody (Santa Cruz Biotechnology, Inc.) for 2 h at room temperature. The results were visualized using an enhanced chemiluminescence detection system (Bio-Rad Laboratories, Inc.), and the density of each band was analyzed using ImageJ software (National Institutes of Health).

For cell cycle analysis, ~2×10⁶ cells were harvested and fixed overnight with 70% ethanol at −20°C. After washing twice with PBS, the cells were suspended in clean PBS with 50 μg/ml propidium iodide and 10 μg/ml RNase A for 30 min in the dark at room temperature. The cells were then analyzed using FACStar Flow Cytometry (BD Biosciences), and the cell cycle distribution was analyzed.

Five-week-old Balb/c female nude mice (Nu/Nu) were obtained from the Laboratory Animal Center of Nantong University. The animal experimental protocol was approved by the Institute of Animal Care and Use Committee of Nantong University (Nantong, China). Fresh surgical tumor tissues (F0) were collected immediately after surgery and cut into 2-3 mm³-sized pieces in DMEM supplemented with penicillin-streptomycin. Tumor fragments were implanted into the right armpit of the mice. When the tumor size reached 100-200 mm³, the samples (referred to as F1) were cut into pieces for passaging in vivo to create F2 xenograft tumors. When the F2 tumor size reached 100-200 mm³, samples were collected and cut into 2-3-mm³ pieces and implanted into the right armpit of mice to create F3. When the F3 tumor size had reached 10-20 mm³, the mice were randomly divided into three groups (n=4 mice/group) and treated through the tail vein with different solutions as follows: The normal saline (NS) group received 100 μl saline solution; the negative control (NC) group received d pLKD-U6-shRNA lentivirus (1×10⁸ pfu) in 100 μl saline; and the last group received pLKD-U6-MSI1-shRNA lentivirus (1×10⁸ pfu) in 100 μl saline. Seven days later, lentivirus administration was repeated. The tumor size and growth rate were monitored and measured using a caliper every 5 days. The approximate tumor volume (V) was calculated using the following equation: V=(longest diameter x shortest diameter²)/2.

The results are shown as the mean ± standard deviation. The data were analyzed using the Duncan test following an analysis of variance for multiple comparisons. Differences were considered statistically significant when P<0.05.

To determine MSI1 expression in OSCC, RT-qPCR analysis was conducted. The results demonstrated that the expression of MSI1 in OSCC tissues was markedly higher compared with that in adjacent healthy tissues (Fig. 1A). In addition, MSI1 expression in the OSCC cell lines was increased compared with that in normal tongue epithelial tissues, and was highest in HSC-3 followed by Ca9-22 cells. However, as Ca9-22 cells have been contaminated with the MSK-922 cell line, which is of head and neck SCC origin, HSC-3 cells for further experiments. (Fig. 1B-D). These data indicated that MSI1 may contribute to OSCC progression, but the underlying mechanism requires further investigation.

To assess the effect of MSI1 on the proliferation and apoptosis of OSCC cells, we established HSC-3 cells with either MSI1 suppression or MSI1 overexpression plasmids. MSI1 expression increased in the MSI1 group, but decreased in the sh-MSI1 group, at both the mRNA and protein levels in HSC cells (Fig. 2A and B). The CCK-8 assay then revealed that MSI1 silencing inhibited cell proliferation compared with control cells (Fig. 2C). In addition, the ratio of BrdU-positive cells among HSC-3 cells with MSI1 suppression was significantly increased compared with that in the control cells (Fig. 2D and E). The results also demonstrated that MSI1 suppression in OSCC cells resulted in markedly lower invasive and migrative ability compared with control cells (Fig. 3).

To further investigate the effect of MSI1 on tumor formation, human OSCC tissues were transplanted into nude mice. The mice were then injected through the tail vein with saline solution (NS), negative control (NC) or MSI1 suppression viral vector. The increase in tumor volume was measured every 5 days. As shown in Fig. 4, MSI1 silencing inhibited tumor growth compared with that in the NS and NC groups. In addition, tumor weight was lower in the sh-MSI1 group compared with that in the NS or NC groups. Western blot analysis revealed that the levels of MSI1 significantly decreased following injection with the sh-MSI1 virus. These data indicate that MSI1 knockdown inhibited the tumor-forming ability of OSCC cells in vivo.

Cell proliferation is directly associated with modulation of the cell cycle; therefore, the cell cycle distribution was then analyzed using flow cytometry. As shown in Fig. 5A and B, the proportion of HSC-3 cells with MSI1 suppression in the S phase was markedly higher compared with control cells.

To explore the molecular mechanism by which MSI1 regulates the OSCC cell cycle, the expression levels of c-Myc, cyclin D1, p21 and p27 were determined. Western blot analysis revealed that the expression of both c-Myc and cyclin D1 decreased in cells with MSI1 suppression; therefore, it was hypothesized that MSI1 may cause cell cycle arrest in part by inhibiting the expression of the c-Myc and cyclin D1 proteins (Fig. 6).

In addition, western blot analysis revealed that the expression of p21 and p27 was upregulated in OSCC cells exhibiting MSI1 suppression, which is consistent with previously reported results (21,22). Furthermore, predictive software (e.g., STRING, StarBase) indicated that c-Myc is also a target downstream protein of MSI1, therefore, luciferase vectors containing wild-type and mutant MSI1-binding sequences were next constructed. The results revealed that the relative luciferase activity in cells with MSI1 suppression was significantly increased compared with that in the controls (Fig. 5C); however, no significant difference was observed between the MSI1 suppression and control groups in cells transfected by c-Myc mutant suppression vector. These results indicated that MSI1 suppression promoted cell cycle arrest, in part by binding to c-Myc.

A large body of evidence has shown that STAT3 expression plays a key role in cancer cell survival, growth and invasion. As shown in Fig. 7, activation of STAT3 at Ser-727 was increased in OSCC tissues compared with that in adjacent healthy tissues, whereas STAT3 at Ser-727 was inhibited in HSC-3 cells following transfection with MSI1-silencing plasmids. These results indicated that MSI1 suppressed OSCC growth in part by inhibiting the activation of STAT3 signaling.