Reduced NKX2.1 Expression Predicts Poor Prognosis of Gastric Carcinoma

Thyroid transcription factor-1 (NKX2.1/TITF-1) is a member of the thyroid tissue-specific transcription factor family that has been proven to be closely associated with many human diseases. Recently, it was reported that NKX2.1 expression is lost or reduced in some human cancers such as lung cancer and thyroid cancer. However, there was insufficient data to suggest that NKX2.1 functionality could be used as a prognostic factor. Therefore, this study aims to investigate NKX2.1 expression and its prognostic significance in primary gastric carcinoma. Then, we attempted to investigate if NKX2.1 expression was related to the clinicopathological characteristics and prognosis of gastric carcinoma (GC)patients. The expression levels of NKX2.1 were analyzed in tissue samples from 205 gastric carcinoma patients by real-time quantitative PCR (qRT-PCR), Western blotting, and immunohistochemical staining(IHC). Our qRT-PCR results showed that the expression of NKX2.1 mRNA was reduced in tumor tissue samples compared with that in matched adjacent non-tumor tissue samples (P<0.001); this finding was confirmed by Western blot analysis (P<0.001). Our immunohistochemical staining data indicated that NKX2.1 expression was significantly decreased in 87 of 205 (42.4%) gastric carcinoma cases. Kaplan-Meier survival curves revealed that the decreased expression of NKX2.1 was significantly associated with poor prognosis in gastric carcinoma patients (P<0.001). Multivariate Cox analysis identified NKX2.1 expression as an independent prognostic factor for overall survival (P = 0.005). Furthermore, the functions of Nkx2.1 were analyzed with respect to the proliferation, migration, and invasion of GC cell lines. Our data suggest that NKX2.1 may function as a tumor suppressor in primary gastric carcinoma and that its reduced expression independently predicts an unsatisfactory prognosis in gastric carcinoma patients.


Introduction
Gastric cancer is one of the most common malignant tumor types worldwide [1] . According to global cancer statistics, gastric cancer (GC) is the fourth most common malignant tumor type and the second most common cause of cancerrelated deaths (of which it accounts for 11%), with approximately 1,000,000 new cases and 800,000 deaths occurring in 2011 [2,3]. Treatment of gastric cancer previously consisted of surgery, chemotherapy, and radiation therapy, but there were no effective targeted therapies [4] . Therefore, many researchers are investigating the molecular markers of GC to uncover new therapeutic targets [5], such as Her-2, which has been proven to be positively expressed in some GC patients and associated with disease prognosis and is being studied in the ToGA trial [6]. It has been proven that GC tumor progression is associated with a multistep process involving the activation of oncogenes and the inactivation of tumor suppressor genes [7] . Thus, we seek to better understand the molecular mechanisms of cancer progression and to develop new therapeutic tools based on these mechanisms.
Thyroid transcription factor-1 (NKX2.1/TTF-1), which was first discovered in thyroid tissue, functions as a tissue-specific upstream transcription factor. NKX2.1 mRNA has a restricted tissue distribution [8] . It is expressed at high levels in thyroid and lung tissues and at low levels in other normal tissues such as the forebrain and pituitary gland [9] . Previous studies mainly focused on the association between NKX2.1 and the differentiation of normal tissues. However, researchers recently began to investigate NKX2.1 expression in human tumor tissues and found reduced NKX2.1 expression in some types of cancer, including lung cancer, thyroid cancer and ovarian carcinomas [10][11][12][13][14][15][16][17], which suggested that NKX2.1 may be a putative new tumor suppressor in multiple types of human cancers. However, to the best of our knowledge, there were no previous reports concerning the expression status and prognostic value of NKX2.1 in primary gastric carcinoma.
In our study, the expression of NKX2.1 in primary gastric carcinoma was assessed using quantitative real-time PCR (qRT-PCR), western blotting and immunohistochemistry. Additionally, we identified the relationship between NKX2.1 expression and clinicopathological features, and we evaluated its prognostic value with respect to post-resection survival in gastric cancer.

Patients and Samples
Clinicopathological data from 205 gastric cancer patients who underwent surgical resection at Sun Yat-sen University Cancer Center from January 2003 to December 2006 were retrospectively analyzed. Patients who met the following eligibility criteria were included: (1) diagnosis of gastric carcinoma identified by histopathological examination; (2) surgical history that included gastrectomy and lymphadenectomy; (3) availability of complete follow-up data; (4) no preoperative treatment, such as chemotherapy and radiotherapy; (5) no history of familial malignancy or other synchronous malignancy (such as GIST, esophageal cancer, or colorectal cancer); (6) no recurrent or remnant gastric cancer; and (7) no death during the perioperative period. The tumor resection and D2 lymphadenectomy were performed by experienced surgeons, and the surgical procedures, which followed the Japanese Gastric Cancer Association (JGCA) guidelines, were similar in all patients who underwent radical resections.
Fresh gastric cancer and adjacent non-tumor tissue samples (n531) were obtained from 31 gastric cancer patients who underwent surgical resection at the Sun Yat-sen University Cancer Center between 2009 and 2011. After surgical resection, the fresh tissue samples were immediately immersed in RNAlater (Ambion, Inc., USA) and stored at 4˚C overnight to allow thorough penetration of the tissues; the samples were then frozen at 280˚C until RNA extraction. Both the tumor tissues and the adjacent non-tumor tissues, which were located more than 3 cm away from the gastric cancer, were sampled and verified by pathological examination. Paraffin-embedded samples were obtained from the 205 gastric cancer patients who underwent surgical resection at the Sun Yat-sen University Cancer Center between 2003 and 2006. Each tumor sample was assigned a histological grade based on the World Health Organization (WHO) classification criteria. All the patients were staged using the 7th edition of the International Union against Cancer (UICC) Tumor-Node-Metastasis (TNM) staging system.

Extraction of total RNA and real-time quantitative PCR
Total RNA was extracted using TRIzol (Invitrogen, Carlsbad, California, USA) according to the manufacturer's protocol. The total RNA concentration was assessed by measuring its absorbance at 260 nm using a NanoDrop spectrophotometer (ND-1000, Thermo Scientific, USA). Reverse transcription (RT) to synthesize the first-strand of cDNA was performed with M-MLV reverse transcriptase (Promega, USA) according to the manufacturer's recommendations, using 2 mg of total RNA as a template. To evaluate the relative mRNA levels of NKX2.1 and GAPDH (used as an internal control), the resulting cDNA was subject to real-time quantitative PCR analysis using the following primers: NKX2.1 forward, 5'-GGACGGGAGCTGGGGAGAGG-3', and reverse, 5'-ATTTTCGCGGAGGGCGGTCG-3'; and GAPDH forward, 59-CTCCTCCTGTTCGACAGTCAGC-39, and reverse: 59-CCCAATACGACCAAATCCGTT-39. Gene-specific amplification was performed using an ABI 7900HT real-time PCR system (Life Technologies, Carlsbad, California, USA) with a 15-ml PCR mix containing 0.5 ml of cDNA, 7.5 ml of 2 x SYBR Green master mix (Invitrogen, Carlsbad, California, USA), and 200 nM of the appropriate oligonucleotide primers. The mix was preheated at 95˚C (10 min) and then amplified at 95˚C (30 sec) and 60˚C (1 min) for 45 cycles. The resolution curve was measured at 95˚C for 15 sec, 60˚C for 15 sec and 95˚C for 15 sec. The Ct (threshold cycle) value of each sample was calculated by the instrument's software (SDS 2.3), and the expression levels of NKX2.1 mRNA were normalized to that of GAPDH. The data were analyzed using the comparative threshold cycle (2-DCT) method.

Western blot analysis
The frozen tissue samples from patients with gastric cancer, including the tumor and normal tissue, were homogenized in RIPA lysis buffer, then the lysates were cleared by centrifugation (12,000 rpm) at 4˚C for 15 min. Approximately 40 mg of each protein sample was resolved on a 12% SDS-PAGE gel and transferred to PVDF membranes. After blocking for 60 min with 5% non-fat milk, the membranes were incubated overnight at 4˚C with a primary polyclonal antibody against NKX2.1 (Abcam, USA, diluted 1:1000). The membranes were then washed three times with TBST for 10 min each and probed with an HRP-conjugated secondary antibody (Immunology Consultants Laboratory, USA, diluted 1:2000) for 60 min at room temperature. The membranes were then washed three times with TBST and developed using an enhanced chemiluminescence system (ECL, Pierce).

Immunohistochemistry analysis
The tissue sections were deparaffinized with dimethylbenzene and rehydrated in 100%, 95%, 90%, 80% and 75% ethanol. After three washes in phosphatebuffered saline (PBS), the slides were boiled in antigen retrieval buffer containing 0.01 M sodium citrate-hydrochloric acid (pH56.0) for 15 min in a microwave oven. After rinsing with PBS, the tissue sections were incubated with primary antibody. The slides were then rinsed in 3% peroxidase quenching solution (Invitrogen) to block endogenous peroxidase. The sections were incubated with a mouse monoclonal antibody against NKX2.1 (Abcam, USA, at a 1:500 dilution) at 4˚C overnight, followed by incubation with horseradish peroxidase (HRP) (ChemMateTM DAKO EnVisionTM Detection Kit) at room temperature for 30 min. After washing in PBS, the antigen-antibody complex were visualized using 3, 3'-diaminobenzidine (DAB), and all the slides were counterstained with hematoxylin. Our negative controls were adjacent sections processed as described above but incubated overnight at 4˚C in blocking solution without the primary antibody.

Cell culture and cellular growth rate
The human GC cell lines, AGS, MKN-45, SGC-7901, HGC-27, MPC-803, and the normal liver cell line GES were obtained from the Committee of Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China). All cells were cultured in RPMI 1640 supplemented with 10% fetal bovine serum (FBS) in 5% CO2 at 37˚C.

Recombinant adenovirus and tumor cell infection
The AD-NKX2.1 and AD-GFP (as a negative control) were bought from HanHeng Biology Company in Shanghai Province, China. The recombinant adenoviruse was stored at 280˚C for use. The SGC-7901 cells were cultured in 6well plates and infected with adenovirus (Ad-NKX2.1 and Ad-GFP) at a multiplicity of infection (MOI) of 20. The efficiency of infection was estimated by western-blot Cell proliferation assay Cellular growth curves were plotted using cellular viability values measured by the MTS method (Cell Titer 96 Aqueous One Solution Cell Proliferation Assay solution; Promega). GC cells transfected with gal-siRNA, NC , galectin-3overexpression were seeded at 1000/cells per well in 96-well flat-bottomed microtiter plates in a final volume of 200 mL of culture medium per well. Cells were incubated in a humidified atmosphere (37˚C and 5% CO2) for 24, 48, 72, 96, 120, 144, 168 hours after transfection. MTS assays were used to measure cell proliferation, with 5 replicates at each time point. Statistical analyses were carried out using the two-tailed unpaired Student's t-test.

Cell migration assays
For transwell migration assays, 5.0 6 10 4 gastric cancer cells in 200 ml of serumfree DMEM were added to cell culture inserts with an 8-mm microporous filter without an extracellular matrix coating (Becton Dickinson Labware, Bedford, MA). Fivehundred microliters of DMEM containing 10% FBS was added to the bottom chamber. After 24 h of incubation, the cells on the lower surface of the filter were fixed, stained and examined microscopically. The stained cells were counted under an inverted microscope (10 fields per membrane). Each experiment was performed in triplicate. Statistical analyses were carried out using the two-tailed unpaired Student's t-test.

Matrigel invasion assay
The Matrigel invasion assay was performed in a chamber system consisting polycarbonate membrane inserts with 8-mm pores (Corning, USA) placed in 24well cell culture insert companion plates. The inserts were coated with a thin layer of 0.5 mg/ml Matrigel Basement Membrane Matrix (BD Biosciences, Bedford, MA). The invasion assay was conducted at 48 hours after the cancer cells were infected with Ad-control/Ad-NKX2.1/siNKX2.1. The cells (1610 5 in 200 ml of growth medium without FBS) were placed in the upper chamber, and 0.5 ml of growth medium containing 10% FBS was placed in the lower chamber. The cells were incubated at 37˚C and allowed to invade through the Matrigel layer for 48 hours. After incubation, the insert membranes were fixed with 75% methanol for 30 hours. The cells on the upper surface were removed with cotton-tipped swabs, and the invading cells on the lower surface were stained with 0.5% crystal violet containing 20% methanol for 60 minutes. The stained cells were counted under an inverted microscope (10 fields per membrane). Each experiment was performed in triplicate. Statistical analyses were carried out using the two-tailed unpaired Student's t-test

Statistical analysis
Differences in mRNA and protein expression between the tumor samples and the paired adjacent non-tumor tissue samples were evaluated with a paired-samples ttest. The x2 test was used to analyze the relationships between NKX2.1 expression and various clinicopathological parameters. Survival curves were calculated using the Kaplan-Meier method and compared by the log-rank test. The Cox proportional hazards regression model was used for univariate and multivariate analyses to study the effects of the clinicopathological variables and NKX2.1 expression on survival. The results were expressed as mean ¡ SD and analyzed using the Students't-test. Differences were considered significant at p , 0.05.

NKX2.1 mRNA expression analyzed with qRT-PCR
Transcription of the NKX2.1 gene was assessed by qRT-PCR analysis of 31 pairs of resected specimens (tumor tissue samples and matched adjacent non-tumor tissue samples) from eligible gastric cancer patients. NKX2.1 mRNA levels were significantly reduced in 23 (74.2%) tumor tissue samples relative to the adjacent non-tumor tissue samples (P,0.001, Figure 1).

NKX2.1 expression analyzed by Western blotting
NKX2.1 protein levels in the resected gastric cancer samples were determined by Western blotting. Our results showed an NKX2.1 band at the expected size of 38 kDa, and the amount of NKX2.1 protein present was then measured by densitometry. As shown in Figure 2, when compared with the matched adjacent  Figure 2A and Figure 2B). These findings were consistent with those from the qRT-PCR analysis.
The association between the levels of NKX2.1 expression and clinicopathological characteristics, based on immunohistochemical staining To obtain further insight into the effect and prognostic value of NKX2.1 expression in gastric cancer patients, paraffin-embedded tissue sections (n5205) of histopathologically confirmed gastric carcinoma were examined using immunohistochemistry. NKX2.1 immuno-positivity was significantly different between the tumor tissue samples and the adjacent non-tumor ones. NKX2.1 expression was localized to the cytoplasm in 118 (57.6%) of the tumor tissue samples, whereas the remaining 87 cases (42.4%) displayed low or no cytoplasmic NKX2.1 expression (Table 1). Based on the categories defined above, the expression status of NKX2.1 correlated significantly with age (P50.002), tumor size (P50.004), depth of tumor infiltration (T stage, P ,0.001) and TNM stage (P ,0.001) but not with gender (P50.861), local lymph node metastasis (N stage,

Expression of NKX2.1 and clinical outcome
The 5-year overall survival rates in patients with high and low NKX2.1 expression were 67.5% and 44.3%, respectively. The overall survival of patients with low NKX2.1 expression was significantly shorter than that of patients with high NKX2.1 expression (P,0.001, log-rank test, Figure 4). Univariate Cox regression analyses showed that the depth of tumor infiltration, local lymph node metastasis, distant metastasis, TNM stage, tumor size and NKX2.1 expression were significantly interrelated with overall survival (Table 2). Furthermore, a multivariate Cox regression analysis confirmed that local lymph node metastasis (P50.002), distant metastasis (P,0.001) and NKX2.1 expression (P50.005) could serve as independent predictors of the overall survival of patients with gastric carcinoma ( Table 2).

The role of NKX2.1 in cell proliferation, migration and invasion
To evaluate the effect of NKX2.1 on cell proliferation a NKX2.1 expression vector and a GFP-control vector were transfected into SGC-7901 cell lines. The efficiencies of transfection were detected by western blotting (Figure 5B). A cell growth assay revealed that cell growth rates in NKX2.1-transfected cell lines were slower than GFP-transfected gastric cancer cell lines ( Figure 5C). The overexpression of NKX2.1 dramatically reduced the migration and invasion ability of the cells (Figure 6).

Discussion
Tumor progression arises as a consequence of a series of cellular events that involve but are not limited to the activation of oncogenes and the inactivation of tumor suppressor genes [18]. In the early 1980s, researchers cloned tumor suppressor genes and demonstrated their pivotal role in regulating the differentiation of normal cells [19] . More recently, inactivating mutations of tumor suppressor genes have been frequently detected in human tumor tissues, suggesting that the expression levels of certain tumor suppressor genes could serve as important prognostic factors for the corresponding human cancers.
NKX2.1, also named thyroid transcription factor-1 (TTF-1), was first discovered in normal thyroid tissue [8]. It functions as a tissue-specific transcription factor, regulating the expression of downstream genes and participating in the differentiation and development of normal tissues [20]. Recent studies have shown reduced expression of NKX2.1 in several types of  [10]. Similar studies have been performed for thyroid cancer and gynecological malignancies, and similar results were obtained for papillary thyroid carcinoma and human ovarian cancer [14][15][16][17]. These data suggest that NKX2.1 can be utilized as an independent prognostic factor in human cancers. In our study, the expression of NKX2.1 in primary gastric carcinoma was determined by quantitative real-time PCR (qRT-PCR) and Western blotting. Our qRT-PCR results showed that the expression of NKX2.1 mRNA was reduced in tumor tissue samples compared with that in matched adjacent non-tumor tissue samples (P,0.001); this finding was confirmed by Western blot analysis (P,0.001). Thus, it is possible for an inactivating mutation of the NKX2.1 gene to  independent risk factor for overall survival. In our in vitro study, NKX2.1 overexpression diminished the proliferation, migration, invasion of gastric cancer cell lines. Moreover, siNKX2.1 transfected cells showed stronger ability of proliferation, migration and invasion. Consistent with the idea proposed in previous studies that NKX2.1 may function as a prognostic factor for some types of cancer, our conclusion supports the hypothesis that NKX2.1 can serve as an independent prognostic factor for gastric cancer patients after surgery. These data suggest that  published an article in the journal Nature, in which they investigated the probable mechanism of NKX2.1 as a tumor suppressor. After studying NKX2.1 gene function in a mouse model of lung carcinoma, the researchers proposed that NKX2.1 might regulate the differentiation of tumor cells and thus limit their metastatic potential [11] . Recent reports clarified that the reduced expression of NKX2.1 is associated with the dedifferentiation of tumor cells and an increase in their metastatic capacity. However, earlier studies found NKX2.1 to be overexpressed in human lung carcinoma, so NKX2.1 was once surmised to be an oncogene [20,21]. Whether NKX2.1 is an oncogene or tumor suppressor remains a hot debate in the field of cancer research. Experts who support the tumor suppressor gene hypothesis demonstrated that NKX2.1 could decrease the expression of occludin, which is a group of proteins localized to and regulating intercellular junctions, thus inhibiting the metastasis of malignant cells [22]. Another hypothesis proposes that NKX2.1 negatively regulates the expression of myosin binding protein H and increases the migratory ability of tumor cells, thereby functioning as a cancer-promoting gene [12]. Therefore, further investigation into the molecular mechanism of NKX2.1 is anticipated.
In summary, our data suggest that NKX2.1 may function as a tumor suppressor in primary gastric carcinoma and that its reduced expression independently predicts an unsatisfactory prognosis in gastric carcinoma patients. We anticipate that NKX2.1 will serve as a valuable prognostic marker in the future.