A total of 100 NSCLC 10 adjacent non-cancerous tissues (defined as tissue which is at least 3 cm from cancerous region; samples were collected from 100 patients (71 men and 29 women) during surgery at Zhejiang Cancer Hospital (Zhejiang, China) between January 2009 and March 2011. All patients gave written informed consent to participate in the study and to allow their samples to be biologically analyzed. The age of the patients with NSCLC enrolled in the present study ranged from 39 to 76 years (mean, 61.3 years). The exclusion criteria included patients with other types of cancer and those who had received preoperative chemoradiotherapy. Control samples were obtained from the same patient at a site at least 3 cm from the tumor and were approved as control samples by a pathologist. Tissue sections were fixed with 10% formalin for at least 24 h at room temperature and subsequently embedded in paraffin for immunohistochemistry. The tumors were staged according to the pathological tumor/node/metastasis (pTNM) classification (7th edition) of the International Union against Cancer (16). All procedures for sample collection and processing were ratified by the International Review Board of Zhejiang Cancer Hospital (Hangzhou, China; approval no. IRB-2016-134).

Tyrphostin AG490 was purchased from Sigma-Aldrich (Merck KGaA). The Cell Counting Kit-8 (CCK-8) was purchased from Dojindo Molecular Technologies, Inc. The Biotin-Streptavidin HRP Detection kit was purchased from OriGene Technologies, Inc. (cat. no. SP-9000). SUMO4-specific small interfering (si)RNA (forward, 5′-GGAUGGUUCUGUGGUGCAGTT-3′ and reverse, 5′-CUGCACCACAGAACCAUCCTT-3′) and negative control siRNA (forward, 5′-UUCUCCGAACGUGUCACGUTT-3′ and reverse, 5′-ACGUGACACGUUCGGAGAATT-3′) were designed and synthesized by Shanghai GenePharma Co., Ltd. Cisplatin was purchased from Jiangsu Hanson Pharmaceutical Co. (http://www.hansoh.cn). The SUMO4 antibody was purchased from Abcam (cat. no. ab126606), phosphorylated (p-)JAK2 and p-STAT3 antibodies were from Cell Signaling Technology Inc. (cat. nos. 3771 and 9145), antibodies against vimentin, E-cadherin, N-cadherin, JAK2 and STAT3 were purchased from ProteinTech Group, Inc. (cat. nos. 10377-1-AP, 20874-1-AP, 22018-1-AP, 17670-1-AP and 10253-1-AP).

Human NSCLC cell lines A549, NCI-H1650 and SK-MES-1 were purchased from The Cell Bank of Type Culture Collection of the Chinese Academy of Sciences. All cell lines were cultured in DMEM (Cytvia) supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin (Transgen Biotech Co., Ltd.) in humidified air at 37°C (5% CO₂).

Cells were seeded into 6-well plates. At 60–70% confluence (5×10⁵ cells/well), SUMO4 siRNA (100 nM) and scrambled control siRNA (100 nM) were transiently transfected into the cells using Lipofectamine^® 2000 Reagent (Invitrogen; Thermo Fisher Scientific, Inc.) according to manufacturer's instructions. The transfection reagent was removed after 5 h, and the cells were harvested after 48 h.

IHC staining of tumor tissues was performed on 5-µm sections. The general procedure for IHC staining was performed as previously described (17). The sections were blocked with blocking serum from the ABC Vectastain kit (cat. no. PK-6100; Vector Laboratories, Inc.) at room temperature for 30 min, followed by incubation with primary antibody against SUMO4 (1:50; cat. no. ab126606; Abcam) overnight at 4°C. Then the sections were incubated with a horseradish peroxidase-conjugated mouse anti-rabbit Ig antibody at room temperature for 1 h, followed by staining with the chromogen diaminobenzidine (Zhongshan, Beijing, People's Republic of China) until a brown color was shown. The slides were counterstained with Mayer's hematoxylin at room temperature for 10 min. Random fields from each slide were viewed under a light microscope (Olympus DP73; Olympus Corporation) at ×20 magnification.

Proteins extracted from cells were denatured in RIPA buffer (150 mM NaCl, 0.1% SDS, 25 mM Tris-HCl pH 7.6, 1% sodium deoxycholate and 1% NP-40) with protease inhibitors. The concentrations of protein were detected using a BCA protein assay kit (cat. no. P0010S; Beyotime Institute of Biotechnology). Equal amounts of total protein (50 µg) were analyzed by SDS-PAGE. Proteins were separated on a 12% polyacrylamide gel and transferred to a nitrocellulose membrane. The membrane was blocked for 1 h at room temperature using blocking buffer (0.5% fat-free milk). After blocking, the blot was probed with primary antibodies (dilution, 1:1,000) at 4°C overnight. Subsequently, the blot was incubated with HRP-labeled secondary antibodies diluted in PBS buffer (1:2,000; cat. no. SA00001-2; ProteinTech Group, Inc.). ECL reagent was used for visualization. The images were obtained using the Bio-Rad Imaging System. Signal quantification was obtained using Quantity One software (version 4.6.6; Bio-Rad Laboratories, Inc.) and normalized to GAPDH.

A total of 5×10⁵ H1650, A549 and sk-mes-1 cells were seeded independently into each well of 6-well plates overnight. A scratch was made on the monolayers with a 200-µl pipette tip, and washed with PBS to remove the detached cells. Fresh DMEM/F12 medium without serum was added into each well of the 6-well plate. The wounded areas were observed and imaged using a light microscope (magnification ×100) at 0 and 24 h. The migration results were quantified using ImageJ software (version 1.52; National Institutes of Health).

The trypsinized A549, H1650 and sk-mes-1 cells were washed with PBS and resuspended in serum-free DMEM medium. Then, 200 µl cell suspension (1×10⁵/well) was added to the upper chamber with a 50 µl solidified Matrigel-coated membrane. The lower chamber was filled with 800 µl DMEM supplemented with 10% FBS. After 24 h incubation, the chambers were fixed with 100% methanol for 20 min at room temperature, followed by staining with 0.1% crystal violet for 20 min at room temperature. Images were captured with an Olympus fluorescence microscope (magnification ×100).

Transfected H1650, A549 and sk-mes-1 cells were collected (100 µl) at 24 h post-transfection and seeded into 96-well plates at a density of 3×10³/well independently. Following incubation for 0, 12, 24, 36 and 48 h at 37°C, 10 µl CCK-8 reagent was added into each of the wells, and the cells were incubated at 37°C for 2 h. The absorbance was measured at a wavelength of 450 nm using a microplate reader (Bio-Rad Laboratories, Inc.).

Statistical analysis was performed using SPSS 23.0 software (IBM Corp.). All experiments were performed in triplicate and data are presented as the mean ± standard error of the mean. The association between the expression of SUMO4 and clinicopathological parameters was examined using the χ² test. The overall survival (OS) and disease-free survival (DFS) curves were produced using the Kaplan-Meier method, the difference in survival was determined using the log-rank test. P<0.05 was considered to indicate a statistically significant difference. Unpaired Student's t-test was performed for western blotting and wound-healing data analysis. One-way ANOVA with Tukey's test was performed for invasion assay and phosphorylation analysis. P<0.05 was considered to indicate a statistically significant difference. Numerical data are presented as the mean ± standard deviation.

In the present study, 10 adjacent normal lung tissues and 100 NSCLC tissues were collected for IHC staining, in order to explore the association between SUMO4 expression and NSCLC prognosis. The results revealed that SUMO4 was expressed differently in the cytoplasm of these tissues. The tissues were divided into ‘positive’ and ‘negative’ groups, according to SUMO4 expression (Fig. 1). The expression of SUMO4 in NSCLC tissues were significantly higher compared with the adjacent normal lung tissue (Fig. 1). Quantitative analysis of the IHC staining results indicated that 69% of NSCLC samples were positive for SUMO4, while all adjacent normal lung tissues were negative via IHC analysis (Table I; P=1.6862×10⁻⁵).

In order to demonstrate the association between SUMO4 expression and NSCLC prognosis, the expression of SUMO4 was systematically analyzed relative to the clinicopathological characteristics of NSCLC from 100 patients. Based on the IHC staining results of all samples, the following associations were identified: Sex, tumor type, history of smoking and T stage were significantly associated with the expression of SUMO4 (P<0.05), however, age, body mass index (BMI), hypertension, diabetes, history of drinking, tumor location, tumor size, tumor differentiation, N stage, tumor stage and recurrence were not associated with the expression of SUMO4 (P>0.05) (Table II).

To clarify the associations between SUMO4 and NSCLC prognosis, the OS and DFS curves of these patients were generated using the Kaplan-Meier method. It was identified that positive expression of SUMO4 was significantly associated with short DFS and OS times (Fig. 2), indicating poor prognosis.

To explore whether SUMO4 regulates the migratory and invasive abilities of NSCLC cells, SK-MES-1, NCI-H1650 and A549 cells which expressed SUMO4, were transfected with SUMO4 siRNA and negative control siRNA (Fig. 3A). EMT marker proteins including N-cadherin, E-cadherin and vimentin were examined by western blotting as shown in Fig. 3B. A notable decrease in vimentin and N-cadherin, as well as increase in E-cadherin was observed. Cell migration and invasion were evaluated by wound healing and Transwell assays (Fig. 3C-F). It was found that SUMO4 siRNA significantly inhibited the wound closure and invasion capacities of A549, H1975 and SK-MES-1 cells.

To identify whether SUMO4 expression affects NSCLC chemosensitivity, control and SUMO4-siRNA transfected A549, H1975 and SK-MES-1 cells were treated with cisplatin at various concentrations, and the CCK-8 assay was used to assess the inhibition of cell proliferation rate. It was found that the inhibitory effect on the proliferation of cells increased with increasing concentrations of cisplatin. SUMO4 silencing dose-dependently altered inhibition rate compared with the NC group (Fig. 4).

To explore the underlying molecular mechanism of the augmented invasion by SUMO4 depletion, the JAK2/STAT3 pathway was specifically analyzed in the present study. Using SUMO4 siRNA combined with the application of JAK2 kinase inhibitor AG490, it was found that knocking down SUMO4 slightly decreased JAK2 and STAT3 activity, as reflected by the phosphorylated forms of JAK2 and STAT3 (Fig. 5). In addition, inactivation of JAK2 by a specific inhibitor (AG490) resulted in decreased invasive ability of cells, demonstrated by the transwell invasion assay (Fig. 3D).