SATB2 Overexpression Promotes the Proliferation, Migration and Invasion of Oral Squamous Cell Carcinoma by Up-Regulating NOX4

Background: While atypical expression of special AT-rich sequence-binding protein 2 (SATB2) has been approved associated with tumor progression, metastasis and unfavourable prognosis in various carcinomas. However, in oral squamous cell carcinoma (OSCC), both the expressive state and associated functions of SATB2’s are still undened. Methods: Real-time PCR, western blotting, and immunohistochemistry were used to examine SATB2 expression. In vitro experiments including Flow Cytometry, CCK8 assay, migration assay, wound-healing assay were used to investigate the effects of SATB2 on HN4 cell proliferation, migration and invasion ability. Additionally, an orthotopic implantation assay was performed in nude mice to conrm the effects of SATB2 in vivo. Furthermore, a genome wide siRNA knockdown experiment was performed to explore the potential downstream regulatory mechanism of SATB2 in OSCC. Results: We found that , in clinical samples from a retrospective cohort of 58 OSCC patients, high expression of SATB2 is associated with poor prognosis of OSCC patients. In this study, we investigated SATB2 is highly expressed in OSCC tissues and cell lines ,which can promotes OSCC cells’ proliferation, migration, invasion and tumor growth. Following a genome wide siRNA knockdown experiment, we identied NOX4, a bona de downstream target of SATB2, which can partially suppress OSCC proliferation. Furthermore, NOX4 knockdown inhibits tumorigenicity, which can be rescued partially by ectopic expression of SATB2. Conclusion: Our ndings not only indicate overexpression of SATB2 triggers the proliferative, migratory and invasive mechanisms which are important in the malignant phenotype but also identify NOX4 as the downstream gene for SATB2. These ndings indicate that may play a key role in may Chain Reaction; polyvinylidenediuoride interfering-non-target kinase;


Background
Oral squamous cell carcinoma (OSCC) is one of the most common solid malignancies worldwide, accounting for approximately 90% of all oral and maxillofacial malignancies in both sexes. The presence of regional lymph node metastasis has been considered as a reliable indicator of the prognosis in OSCC patients [1]. Surgery is currently the most effective treatment, especially for initial tumor likely to spread, Which in turn develops local invasiveness and lymph node metastasis. In recent years ,the multidisciplinary sequential therapy strategy (including surgery, radiotherapy, chemotherapy, and biotherapy) has been used to treat OSCC, however, there has been no signi cant improvement in the 5year survival rate of patients, especially in those with neck metastasis or at an advanced pathological stage. [2][3][4] STAB2 (Special AT-rich sequence-bind 2) is a nuclear transcription factor that play a vital role in various biological functions, such as osteoblast differentiation and bone matrix formation. [5][6]. After examining protein expression patterns in human normal and cancer tissues, STAB2 was considered to be a tissue-type speci c protein [7] and other studies on its tumorigenic roles have found that it may act as both a tumor suppressor and promoter. For example, SATB2 is often overexpressed in breast cancer [8], while only expressed at low levels in colorectal cancer [9]. Until now, there have been limited studies on SATB2's role in OSCC, although it has been shown to be highly expressed in advanced HNSCC (Head and Neck Squamous Cell Carcionma) where it can promotes HNSCC cells' chemoresistance and governs HNSCC cell survival [10].
The localization of NOX4 (encoded by a 265 kb gene located on chromosome 11q14.2-q21) is cell-type speci c and can be found in mitochondria, endoplasmic reticulum, nucleus, and focal adhesions [11]. Initially considered to be a kidney-speci c protein [12], NOX4 has been detected in numerous other tissues including blood vessels, heart, liver, and neurons [13]. NOX4 has a regulatory role in various cellular functions such as angiotensin II-induced vascularization [14] and insulin-triggered glucose uptake [15].
Abnormal NOX4 expression is associated with a wide range of cancer types including pancreatic and gastric cancer [16,17], melanoma and von-Hippel-Lindau-de cient renal cell carcinoma [18]. NOX4 has been reported as being involved in almost every process of tumor development. NOX4 generate ROS may in uence OSCC tumorigenicity and provide novel observations on metastasis, invasion, DNA damage, epithelial-to-mesenchymal transition (EMT) as well as helping cancer cells develop resistance to chemotherapeutic agents and radiation [11]. In osteosarcoma cell lines, SATB2 enhances migration and invasion by regulating genes involved in cytoskeletal organization. Microarray analysis identi cation of genes differentially regulated by SATB2 included NOX4, which is up-regulated in sh-SATB2 cells [19]. In this study, we tried to clarify the potential molecular mechanisms that SATB2 promotes the proliferation, migration, and invasion of OSCC by targeting NOX4.
In this, we examined STAB2 expression in OSCC samples and cell lines and its clinicopathological signi cance, its biological roles in OSCC cell lines, and the proliferative capacities of SATB2-modi ed OSCC cell lines.

Ethics Statement
The protocol for this study was approved by the Institutional Review Board of Nanjing Medical University and all experiments were performed after obtaining written informed consent for OSCC clinical specimens.

Patient and Tissue Samples
Specimens were obtained from a retrospective cohort of 58 primary human OSCC cases seen from Jan. 2011 to Dec. 2014. 11 samples of normal oral mucosa tissues from other non-cancer surgeries during the same period were collected from patients who underwent surgical resection at the department of oral and maxillofacial surgery, Nanjing Medical University (Nanjing, China). These OSCC patients did not receive any preoperative treatment before surgery. 6 pairs of fresh clinical specimens (for use in RT-PCR and Western blot) were collected from OSCC patients at the same department. Histological examination was performed by senior oral pathologists (according to the diagnosis by Chief Dr Song XL, the senior pathologist at the A liated Stomatological Hospital of Nanjing Medical University), and diagnosis made based on carcinoma cell features as seen under a microscope. Tumors were classi ed according to the International Union Against Cancer (UICC) tumor staging system. All fresh tissue samples were collected and immediately stored at -80 °C until further use.

Mice
Animal studies were performed in an Association for Assessment and Accreditation of Laboratory Animal Care International accredited SPF animal facility and all protocols approved by the Animal Care and Use Committee of the Animal Research Center, Nanjing Medical University.

Cell lines and culture
Human oral keratin cells (OKC) and three human HNSCC cell lines: HN4, HN6, SCC25 were purchased from American Type Culture Collection (ATCC). Cancerous cell lines were cultured in DMEM media (Invitrogen) supplemented with 10% FBS (Hyclone) and 100 units/ml penicillin and streptomycin (Sigma) in 5% CO 2 at 37 °C.

Immunohistochemistry
IHC staining was performed using the standard streptavidin-biotin-peroxidase complex method according to the manufacturers' protocol (Abcam, USA). Brie y, 4 mm tissue sections from representative para n blocks were depara nized in xylene and rehydrated using an ethanol gradient. Endogenous peroxidases were blocked with 3% hydrogen peroxide. Slides were heated for antigen retrieval in 10 mM sodium citrate retrieval buffer (0.01 M sodium citrate and 0.01 M citric acid, pH 6.0), at 95℃ for 5 minutes. Sections were blocked in 10% normal swine serum for 20 minutes and incubated with monoclonal rabbit primary antibody (anti-SATB2, 1:200 dilution; Abcam, USA), monoclonal rabbit primary antibody (anti-NOX4, 1:400 dilution; Abcam, USA), at 4 ℃ overnight. Specimens incubated with PBS instead of primary antibodies were used as negative controls. Sections were then incubated with secondary antibodies for 45 minutes at room temperature. Reaction products were developed by 3, 3'diaminobenzidine solution with hydrogen peroxide, followed by hematoxylin counterstaining.

RNA extraction and quantitative real-time PCR
Total RNA was extracted using TRIZOL Reagent (Invitrogen) and cDNA synthesized using a reverse transcription Polymerase Chain Reaction (RT-PCR) Kit (TaKaRa) according to the manufacturer's instructions. Quantitative real-time Polymerase Chain Reaction (qRT-PCR) was performed using SYBR Premix Ex TaqTM II PCR Kit (TaKaRa) and an ABI 7500HT PCR sequencer (Applied Bio-systems). Primers speci city was veri ed by dissociation curve analysis. Data was analyzed using ABI SDS v2.4 software (Applied Biosystems). All qRT-PCR reactions were performed in triplicate. The housekeeping gene GAPDH was used as an internal control.

Western blot analysis
Protein samples were extracted by RIPA (Beyotime, Shanghai, China). Equal amounts of protein lysate were separated by SDS-PAGE, transferred to a polyvinylidenedi uoride (PVDF) membrane (Millipore), then blocked with 3% BSA in Tris-buffered saline with 0.1% Tween 20 (TBS-T) for 1 h at room temperature. The blocked membrane was then incubated with polyclonal primary antibody against (SATB2/MOX4) overnight at 4 ℃, washed three times in TBS-T, then incubated with horseradish peroxidase (HRP)-conjugated secondary antibody for 1 h at room temperature. Antibody-labelled proteins were detected using an ECL western blot detection kit (Bio-Rad Laboratories, Hercules, CA, USA) and X-ray lm (Kodar). Tubulin or GAPDH were used as loading controls.

Immuno uorescence on HN4 cells
Cells were seeded onto 12-mm coverslips in 24 well plates. After 24 h incubation cells were washed in cold PBS, xed with 4% paraformaldehyde for 30 minutes at room temperature, then washed three times with cold PBS. Cells were blocked with 3% BSA and 0.1% Triton-100 for 1 h at 37°C, incubated with a SATB2 primary antibody (Abcam,1:200) overnight at 4 ℃, washed three times with cold PBS, and incubated with FITC-conjugated secondary antibodies for 1 h. Cells were subsequently stained with DAPI for 10 min at room temperature (Roche Diagnostics), sealed with 70% glycerin, and examined using a Nikon uorescence microscope.

Lentivirus generation and cell transfection
Lentivirus encoding full-length human SATB2 cDNA was purchased from Shanghai Cyagen biotechnology Co. Ltd. and used to transduct HN4 and HN6 cell lines which had reached 50% con uency. These cells were grown in DMEM medium (Invitrogen) with 10% FBS (Hyclone) for another 48h, after which stable SATB2 -expressing clones were screened by adding 10 μg/ml puromycin (Sigma) and growing for a further 24 hours. This was repeated twice to acquire Lv-SATB2-HN4 cells and N.Ctransducted cells were used as controls.

siRNA Synthesis and Transfection
Speci c small interfering RNAs (siRNA) against SATB2 and NOX4 were purchased from Shanghai Genepharma biotechnology Co. Ltd. One day before transfection, 2×10 5 HN4 cells (in 2ml antibiotic-free growth medium) were seeded at per well in a 6-well plate (Costar) . The second day , each of 5μL siRNA vector (100ng/μL) or 3μL Lipofectamine TM 2000 (Invitrogen) were diluted in 50μL serum-free Opti-MEM medium (GIBCO,USA), mixed gently and incubated at room temperature for 5 min. Then, the diluted siRNA vector and Lipofectamine TM 2000 were combined, gently mixed, and incubated at room temperature for 20 minutes. During this incubation time, the seeded HN4 cells were washed twice with PBS and 1.9ml serum-free medium was added to each well. 100μL of the siRNA-Lipofectamine TM 2000 mixture was added to each well and gently mixed by rocking the plate. The cells were grown in 5% CO 2 and 95% air at 37℃ for 6 hours and the serum-free medium was exchanged for serum containing medium. Cells were grown for a further 72 h then collected for assays. The siRNA sequences are listed in Supplementary  Table 4.

Flow Cytometry
HN4 and Lv-SATB2 -HN4 cells were trypsinized and resuspended in PBS. For cell cycle analysis, cells were washed in PBS for 3 times, xed in 70% ethanol, and then stained with propidium iodide following RNase treatment. The DNA content and cell cycle distributions were analyzed using FACS ow cytometry (BD, USA) and Cell Quest software (BD Biosciences, USA).

CCK8 assay
Cell proliferation and viability were assessed using a CCK8 assay. 1x10 3 cells were seeded in the wells of a 96-well plate and incubated overnight at 37°C. 10μl CCK8 was mixed with 90μl DMEM medium (Invitrogen) containing 10% FBS (Hyclone), and added to each well and plates were incubated for another 2-4 h. These steps were conducted in a dark environment. After 2-4 hours' incubation, we restarted time points (0, 24, 48 and 72 h), the absorbance was measured at 450 nm using an automatic enzyme-linked immune sorbent assay reader (Molecular Devices, San José, CA, USA).

Cell Migration and Invasion Assays
In wound healing experiments, cells were seeded at 70% con uency in 6-well plates (Costar) lled with DMEM/F -12 containing 10% FBS. 24 hours after seeding, scraped a wound line with a P1000 pipette tip among the con uent monolayers, washed to remove cell debris, and media was replaced. At speci c time points, cells were xed in 3.7% paraformaldehyde and photographed under a phase-contrast microscope.
Migration assays were performed using a trans-well chamber (Corning, NY, USA). 1×10 4 cells (in 200 µL DMEM/F-12) were placed on the upper layer of a cell-permeable membrane and 500 µL DMEM/F-12 containing 10% FBS was placed in the lower chamber. Following an incubation period, the cells that had migrated through the membrane were stained. Matrigel matrix (BD, Billerica, MAUSA) was used to simulate a human basement membrane for the invasion assay. DMEM and Matrigel matrix were mixed and the membrane was also coated with matrigel (1:9 dilution ratio) for testing invasion.

OSCC tumor xenograft formation in nude mice
All animal protocols used in this study were in accordance with the institutional animal welfare guidelines of Nanjing Medical University. Mice were randomly assigned into four groups, each containing six 4-to-6week-old male nude mice. For HN4 cell line, each experimental group was subcutaneously injected with a total of 2 × 10 6 cells of either; HN4, Lv-SATB2-HN4, Lv-SATB2-HN4 with Negative Control, or Lv-SATB2-HN4 with NOX4 knockdown. As for HN6 cell line, each of 5 × 10 6 HN6 and Lv-SATB2-HN6 cells were subcutaneously injected into the left and right anks respectively. Tumor sizes and weight were recorded every three days. All extant mice were euthanized at 4 or 8 weeks after injection then opened at injection sites to con rm tumor size and weight. Tumor volumes were measured by caliper and calculated as follows: Volume (mm 3 ) = D × d 2 × 0.5, where D is the longest, and d the shortest, diameter of the tumor.

Statistical analysis
All statistical analysis were performed using Graph Pad Prism 5.01 (La Jolla, CA, USA) or SPSS 18.0 (Armonk, NY, USA). Pearson χ 2 test was used to analyze the association between SATB2 expression and clinical pathology parameters. Survival rate analysis was analyzed by Kaplan-Meier plot and log-rank test. Independent Student t-test and ANOVA with post hoc test were used for most other analyses as indicated in gure legends. The data are presented as the mean ±S.D. of at least three independent experiments. The P-values were de ned as *P 0.05, **P 0.001 and ***P 0.0001.

SATB2 is overexpressed in OSCC samples and cell lines
Previous studies indicated that SATB2 could be a crucial oncogenic gene during cancer development and progression [7,9]. To examine SATB2 expression in primary OSCC, we rst assessed the expression of SATB2 protein in the clinical specimens retrieved from 58 primary OSCC patients by immunohistochemical staining. Representative immunohistochemical staining of primary OSCC and normal oral mucosa is shown in Fig. 1A. SATB2 protein abundance in these primary OSCC (n = 58) and normal oral mucosa specimens (n = 11) was categorized according to our immunohistochemistry scoring regime. SATB2 levels in the OSCC specimens were graded as negative (0), low (18) and high (40) expression and in the normal samples were graded as negative (9), low (2) and high (0) ( Table 1), indicating that it was overexpressed in the majority of clinical samples. SATB2 was highly expressed in the majority of OSCC specimens (31 of 42) with advanced pathological staging ( -), whereas only a minority (7 of 16) of lower-stage ( -) tumors showed elevated expression. (Fig. 1B, P 0.05).
To support these observations, mRNA from six pairs of fresh oral carcinoma tissues including tongue, gingiva, palate, and the para-carcinoma tissues of the corresponding sites were analyzed by RT-PCR. The mRNA levels of SATB2 were signi cantly increased in tumor group compared to the control (Fig. 1C).
Western blots for SATB2 in the parallel samples also showed the increased expression in the tumor group compared to the control (Fig. 1D). We then assessed the expression pro le of SATB2 in the human tongue (SCC25) and head and neck squamous cell lines (HN4 and HN6), with human oral keratin cells (OKC) as the control and showed that SATB2 mRNA levels were higher in HN4, HN6, SCC25 compared to the OKC control (Fig. 1E). This result was further supported by western blotting using the HN4 cell line (Fig. 1F).

SATB2 overexpression is associated with tumor size, cervical node metastasis, clinical stage, and poor prognosis of OSCC
The epidemiologic information and relevant clinicopathological features of patients are summarized in Table 2. In brief, 37 male and 21 female patients were enrolled with mean age 56.5 years with a range of 39-79 years. The average follow-up period was 54.7 months. There were no signi cant correlations between SATB2 expression with patients' gender, age, and pathological grade, however, a Pearson χ 2 test revealed signi cant associations between SATB2 abundance and tumor size (P = 0.007), cervical nodes metastasis (P = 0.043) and clinical stage (P = 0.031). To examine any potential prognostic value in OSCC, we examined SATB2 expression and patient survival. At last follow-up with the 58 patient cohort, 35 (60.3%) were alive and disease-free, 6 (10.3%) were alive but had recurrence and/or cervical nodal metastasis, and 17 (29.4%) had died due to either a local recurrence, metastasis, or other unrelated reasons. The results from the Kaplan-Meier survival analyses indicated that high SATB2 expression had an adverse prognostic impact on patients' outcomes and was negatively related to overall survival ( Fig. 1G, Log-rank, P = 0.0179).

Establishment of Lv-SATB2-HN4 and Lv-SATB2-HN6 cell lines
HN4 cells were transducted with the full-length human SATB2 cDNA to evaluate its function. Cells were screened twice with puromycin for 72 h post-transduction to ensure stable transduction had occurred. Green uorescent protein was observed by uorescence microscopy 48 h post-transduction (Fig 2A). Both SATB2 mRNA and protein levels were signi cantly increased in Lv-SATB2-HN4 and Lv-SATB2-HN6 cells compared to HN4 and HN6 cells. (Fig 2B). Immuno uorescence indicated that there was stronger immunocytochemical staining in Lv-SATB2-HN4 cells (Fig 2C). These results con rmed that a SATB2 modi ed-HN4 cell line was successfully established and ready for use in subsequent experiments.

SATB2 overexpression promotes HN4 and HN6 proliferation
To further investigate the biological role of SATB2 we used ow cytometry to examine the cell cycle pro le of HN4 before and after SATB2 overexpression. The results showed that compared to HN4 cells, Lv-SATB2 cells had a signi cant decrease in the percentage of cells in G1 phase (HN4:63.33% ±2.44; Lv-SATB2: 51.37 % ±2.96), and an increase in the percentage of cells in S phase (HN4: 23.51% ±8.84; Lv-SATB2: 38.26 %±8.49) (Fig 3A). Cell proliferation was examined using a CCK-8 assay at 3 time points (24,48, and 72 h) and was consistent with the cell cycle analysis, in that from 24 h, the number of viable cells was signi cantly higher in Lv-SATB2 than either HN4 and HN6 (Fig 3B). To further clarify SATB2's tumorigenic potential in vivo, HN4, Lv-SATB2-HN4, HN6, and Lv-SATB2-HN6 cells, were subcutaneously injected into the left armpit of the mice. Eight weeks after injection tumors were resected and collected. Tumor volumes and weights were markedly increased in the Lv-SATB2 groups compared to the control groups (Fig 3C, D). The expression of key molecules that regulate the G1/S phase transition in HN4 and Lv-SATB2-HN4 cells was examined. Persistent SATB2 activation has been associated with the promotion of proliferation, anti-apoptosis, invasion, metastasis, angiogenesis, and immune escaping behaviors, and with genes encoding key cancer-promoting in ammatory mediators in most malignant tumor cells [20].
Interleukin-6 (IL-6), signal transducers and activators of transcription 3 (STAT3), and p-STAT3 levels in the cytoplasm of HN4 and Lv-SATB2 cells were determined by western blot (Fig 3E). The expression of IL-6 (the most important of STAT3 activators) was signi cantly increased after SATB2 overexpression. SATB2 overexpression in HN4 cells results in phosphorylation of Tyr705 in STAT3, leading to its nuclear translocation. STAT3-regulated genes encode cytokines, such as IL-6, which in turn activate the STAT3 signaling pathway and consequently propagate a feed-forward loop between tumor and immune cells in the tumor microenvironment [20].

SATB2 overexpression promotes HNSCC migration and invasion
Metastasis is a feature of most malignant tumors and the culprit of many cancer-related deaths. Previous reports demonstrated that SATB2 promoted head and neck squamous cells proliferation and their survival in the presence of radiation [10]. To explore the speci c roles of SATB2 in OSCC metastasis, we examined their migration and invasion. SATB2 overexpression signi cantly enhance both cell migration, as evidenced by wound-healing and transwell assays, compared to the control cells (Fig. 4A-C) and invasion in the transwell matrigel invasion assays (Fig. 4D). SATB2 overexpression in HN4 cells promoted tumor lung metastasis in one mouse (Fig. 4E), which provided direct evidence to support the hypothesis that SATB2 contributed to cell invasion.
The epithelial-mesenchymal transition (EMT) has been reported as an important biological process in embryonic development and tumorigenesis [21]. During this process, epithelial tumor cells can obtain mesenchymal phenotypes and promote tumor invasion and metastasis. We detected E-cadherin (epithelial marker) and vimentin (mesenchymal marker) protein levels by western blotting in HN4 and Lv-SATB2-HN4 cells. Our data showed decreased expression of E-cadherin and increased expression of Vimentin in Lv-SATB2-HN4 cells compared to their counterparts (Fig 4. F). Collectively, these results indicate that SATB2 overexpression promotes HNSCC migration and invasion both in vitro and in vivo, as well as facilitates the epithelial-mesenchymal transition of OSCC cells.
3.6 NOX4 is up-regulated after SATB2 overexpression in HN4 cells NOX4 knockdown by siRNAs reduces cell proliferation and inhibits OSCC tumor growth.
To better understand the function of NOX4, three NOX4-speci c siRNAs (si-NOX4) were designed to selectively deplete NOX4. Cells treated with scrambled si-RNA were used as small interfering-non-target controls (si-NC). The results from RT-PCR, western blot, and immuno uorescence assays demonstrated successful inhibition in the si-NOX4 group compared to the si-NC group (Fig. 5D-F). Cell proliferation was detected by a CCK-8 assay after 24h, 48h and 72h. The number of viable cells markedly decreased in the siNOX4-treated cells compared to the control and the differences were statistically signi cant at 72 h ( Fig   5G). Cell cycle distribution among cells was determined by ow cytometry to further investigate NOX4's effect on cell growth. We found that NOX4 down-regulation induced cell cycle arrest at the G1/S checkpoint (Fig 5H, Table 3). These results indicate that NOX4 plays an important role in tumor formation in Lv-SATB2-HN4 cells and can be a positive regulator of tumor growth. To investigate NOX4's in vivo tumorigenic potential , scrambled siRNA and siNOX4-treated cells were subcutaneously engrafted in the left armpit of mice, respectively. To minimize the number of mice used, three NOX4-speci c siRNAstreated cells were equally mixed together as the siNOX4 group. NOX4 knockdown could effectively suppress tumor initiation when compared to the control group (Fig 5I).

Discussion
In our previous study, our study con rmed the inhibition of miR-34a in the invasion, proliferation, and migration of the OSCCs, playing a potential tumor suppressor role with SATB2 as its downstream target [22]. In this study, we further investigated the expression pattern of SATB2 in OSCC. SATB2 overexpression was signi cantly associated with tumor size, cervical node metastasis, and clinical stage, all of which have critical clinicopathological relevance and prognostic signi cance for OSCC patients. We also observed prominent expression of SATB2 in OSCC tissues compared with para-carcinoma oral tissues. SATB2 overexpression in HN4 cells promoted proliferation, migration, and invasion, as well as facilitating EMT and activating the Janus kinase (JAK)/STAT3 signaling pathway. These ndings suggest that SATB2 overexpression in important for progression in oncogenesis, like the role of SATB1 in breast cancer [23]. Because of the limited number of patients enrolled in this study, more patients from multiple institutions are needed to de nitively establish SATB2's overexpression pattern as well as its diagnostic utility in OSCC.
In this study, our ndings support that hypothesis that SATB2 has in uence on the determination of OSCC proliferation, migration, invasion, and tumor growth in vitro and in vivo. In future experiments, we aim to examine the expression of key molecules that regulate the G1/S phase transition in HN4 cells and Lv-SATB2-HN4 cells.
In Lv-SATB2-HN4 cells, high levels of IL-6 and p-STAT3 have been discovered, but total STAT3 expression is stable. Phosphorylated STAT3 is a nuclear transcription factor involved in tumor proliferation, survival, angiogenesis, and invasion, in addition to in uencing genes encoding key cancer-promoting in ammatory mediators [20,24]. IL-6 has been described as acting like a cytokine and activates STAT3 phosphorylation signaling. Obviously, STAT3 activation can be achieved via other signaling pathways, such as receptor-tyrosine kinases, which are dysregulated in cancer. It has been reported that IL-6 promotes head and neck tumor metastasis and EMT via the STAT3/SNAIL signaling pathway [24]. Like SATB1 in breast cancer [23], IL-6 promotes growth and invasion of breast cancer cells through STAT3dependent up-regulation of the NOTCH signaling pathway [20,24,25].
Previous studies have suggested that knocking down, or silencing, NOX4 caused increased cell proliferation in hepatocytes and hepatocarcinoma cells and in vivo, suggesting that it suppresses liver cell proliferation [26]. Our ndings show that NOX4 was prominently expressed in lentiviral transducted SATB2-speci c HN4 cells, and that silencing expression using NOX4-targeted siRNA decreased viability of HN4 cells. In this study, the involvement of NOX4 in proliferation had been con rmed by CCK-8 assay and Flow Cytometry. NOX4 appears to be a constitutively active enzyme that is transcriptionally regulated [11]. As a key oxygen sensor, NOX4-derived H 2 O 2 plays diverse roles in cell proliferation, migration, and death. Increased NOX4 expression has been observed in cancer, which ptomotes in metastasis, angiogenesis, DNA damage, anti-apoptosis and EMT [27].
In future experiments, we will investigate these areas and examine whether there is a direct association between SATB2 and the NOX4 promoter.

Conclusion
Our study shows that OSCC patients with high SATB2 levels usually have poorer prognosis. In both in vitro and in vivo, SATB2 overexpression promotes proliferation, partially via up-regulating NOX4, and enhanced migration and invasion of OSCC cells. Our ndings suggest that SATB2 could not only serve as a novel and viable biomedical diagnostic and prognostic biomarker, but also may be a potential therapeutic target in OSCC. The study which including human and animal approval were acquired from the Ethics Committee of The A liated Stomatology Hospital of Nanjing Medical University .All experiments were performed after obtaining written informed consent for OSCC clinical specimens.

Consent for publication
We assure that the material is original and it has not been published elsewhere yet.

Data availability statement
All datasets presented in this study are included in the article/additional les. Table 2 Association between SATB2 expressions with key clinicopathological parameters in 58 OSCC specimens Table 4