Ang-2 promotes lung cancer metastasis by increasing epithelial-mesenchymal transition

Lung cancer is the most common malignant tumor with increasing angiopoietin-2 (Ang-2) and a high rate of metastasis. However, the mechanism of Ang-2 enhancing tumor proliferation and facilitating metastasis remains to be clarified. In this study, Ang-2 expression and its gene transcription on effects of biological behaviors and epithelial-mesenchymal transition (EMT) were investigated in lung cancers. Total incidence of Ang-2 expression in the cancerous tissues was up to 91.8 % (112 of 122) with significantly higher (χ2=103.753, P2=7.883, P=0.005), differentiation degree (χ2=4.554, P=0.033), tumor node metastasis (TNM) staging (χ2=5.039, P=0.025), and 5-year survival rate (χ2 =11.220, P2=18.881, P2=0.81, P=0.776) or III & IV (χ2=1.845, P=0.174). Over-expression of Ang-2 or Ang-2 mRNA in lung A549 and NCI-H1975 cells were identified among different cell lines. When silencing Ang-2 in A549 cells with specific shRNA-1 transfection, the cell proliferation was significantly inhibited in a time-dependent manner, with up-regulating E-cadherin, down-regulating Vimentin, Twist, and Snail expression, and decreasing invasion and metastasis of cancer cell abilities, suggesting that Ang-2 promote tumor metastasis through increasing EMT, and it could be a potential target for lung cancer therapy.


INTRODUCTION
Lung cancer is one of the most common malignancy tumors and the leading cause of cancer-related death [1,2]. The early symptoms of patients with lung cancer are not obvious, with diagnosis difficult at early stage, tumor metastasis, drug resistance, and poor prognosis [3]. The occurrence and development of lung cancer is related to oncogene activation, tumor suppressor gene inactivation, protein functional effector sncRNAs (pfeRNAs), angiogenesis, and inflammation [4,5]. Increasing evidence implicates that exosomes confer stability and can deliver their cargos such as proteins and nucleic acids to specific cell types, which subsequently serve as important www.impactjournals.com/oncotarget/ Oncotarget, 2018, Vol. 9, (No. 16), pp: 12705-12717 Research Paper messengers and carriers in lung carcinogenesis [6]. However, angiogenesis is a fundamental process involving a variety of pathological processes, sustains the progression of many neoplastic diseases, and successful establishment depends on a complex process of endothelial proliferation and organization [7,8]. Moreover, it may enhance tumor cell proliferation and resistance to apoptosis, and facilitate metastasis. Angiopoietins (Angiopoietin-1, Ang-1 and Angiopoietin-2, Ang-2), Tie2 ligand-receptor, and vascular endothelial growth factor (VEGF)-mediated pathway are essential for the regulation of vascular maturation or stability, and have been implicated in the control of physiological angiogenesis [9,10].
Recently, Ang-2 is an important proangiogenic factor that has been implicated in mediating inflammatory processes, and upregulated in multiple inflammationrelated tumors or signaling pathway [11]. Lung cancer vasculature originates because of angiogenesis, vascular sheath growth, and endothelial progenitor cell growth [12]. Lung tissue hypoxia in the absence of vascularization results in tumor cells to produce lots of angiogenic factors inducing angiogenesis for tumor growth [13,14]. Abnormality of Ang-2 expression has been reported to promote blood vessels and increase vascular permeability in ischemic and/or hypoxic environment [15,16]. With the volume increasing, intratumoral hypoxia increased Ang-2 expression to promote angiogenesis for tumor metastasis [17,18], and Ang-2 and Ang-2 mRNA in tissue or sera have been useful diagnostic biomarkers for lung cancer [19,20]. However, the relationship between Ang-2 expression and proliferation, invasion, migration of cancer cells remains to be explored. Therefore, the objectives of this study were to investigate the alteration Ang-2 expression in lung tissues, and observe on effects of biological behaviors and epithelial-mesenchymal transition (EMT) abilities of lung cancer cells in vitro by silencing Ang-2 with RNA interference technology.

Pulmonary Ang-2 expression and clinicopathological characteristics
Ang-2 expressions in 122 cases with lung cancers and their paracancerous tissues with the immunohistochemical analysis are shown in Figure  1. The Ang-2 expression with deeper brown staining particles in the cancerous tissues ( Figure 1A1 and 1A2) were mainly localized in cytoplasm and cell membranes, and with lighter cytoplasm staining in their paracancerous tissues ( Figure 1B1 and 1B2). Total incidence of Ang-2 expression in the cancerous group was up to 91.8 % (112 of 122) with significantly higher (χ 2 =103.753, P<0.001) than that in the surrounding group (27.9%, 34 of 122). The brown Ang-2 expressions were gradually increasing with clinical staging and showed very strong staining at advanced lung cancer. Correlation between Ang-2 levels and the clinicopathological parameters in cancerous tissues are shown in Table 1

Univariate and multivariate analysis of Ang-2 expression
Univariate and multivariate analysis of high Ang-2 expression in the tissues of lung cancer are shown in Table 2. Significant difference was found between high Ang-2 expression and 5-year survival rate of patients. The univariate analysis showed that Ang-2 expression (P<0.001), gender (P=0.042), differentiation degree (P<0.001), and TNM stage (P<0.001) were potential factors influencing survival of HCC patients. Further, the multivariate analysis showed that Ang-2 expression (P<0.001), differentiation degree (P=0.036), and TNM stage (P=0.030) were independent markers for HCC prognosis ( Table 2). The Kaplan-Meier survival curves showed that the 5-year survival rate of high Ang-2 patients was significantly lower than that of cases with low or no expression (χ 2 =16.486, P<0.001; Figure 2A). However, significant difference was found only at TNM stage I (χ 2 =18.881, P<0.001; Figure 2B) but not at II (χ 2 =0.81, P=0.776; Figure 2C) or III & IV (χ 2 =1.845, P=0.174; Figure 2D), indicated that the high Ang-2 level be an independent factor for poor outcome.

Difference of lung cancer cell Ang-2 expression
The comparative analysis of Ang-2 expressing differences among lung cancer cell lines is shown in Figure  3. Ang-2 expression in lung cancer (SPC-A-1, NCI-1650, A549, and NCI-1975) cells at protein-( Figure 3A and 3B) or mRNA-( Figure 3C) level was significantly higher (1.5~4 times) than that in lung Beas-2B cells, especially in overexpression of A549 and NCI-H1975 cells. Furthermore, Ang-2-shRNAs and a negative-control shRNA (NC-shRNA) were successfully transfected into A549 cells, and fluorescence photomicrographs were observed at 24h ( Figure 3D and 3E); Ang-2 at protein level at 48 h showed the different silencing efficiencies ( Figure 3F) with high efficiency in shRNA-1, low efficiency in shRNA-2 and shRNA-3 ( Figure 3G). Similar interference effects were observed at Ang-2 mRNA level ( Figure 3H), showed that the most effective shRNA-1 could significantly downregulate Ang-2 expression at mRNA or protein level. www.impactjournals.com/oncotarget

Ang-2 affected biological behaviors & EMT
After the A549 cells with the most effective Ang-2-shRNA-1 transfection above, the abilities of cell proliferation, migration, invasion and EMT are shown in Figure 4. After transfected with shRNA-1 for 48 h, the cells showed lower proliferate abilities than those in the NC-or control-group at a time-dependent manner ( Figure  4A and 4B). The abilities of the A549 cell migration and invasion migrated through the membrane in the migration chamber without (the shRNA group, shRNA, Figure  4C1) or with the transwell-precoated Matrigel lower than those in the negative control group (NC, Figure  4C2) or in the blank control group (control, Figure 4C3). The quantitative comparison of down-regulating Ang-2 on effect of invasion and migration of A549 cells are summarized in Table 3. Significant statistical difference ( Figure 4D, migration: t=155.449, P<0.001; Figure 4E, invasiveness: t=142.933, P<0.001) was found between the shRNA-1-and the control-group, and silencing Ang-2 transcription remarkably decreased the abilities of proliferation, migration and invasion of A549 cells.
Above the basis of Ang-2 promoting lung cancer A549 cell migration and metastasis, the alterations of the EMT-related epithelial indicator (E-cadherin), mesenchymal marker (Vimentin, VIM), and transcriptional factors (Snail and Twist) expression that involved in lung cancer development were analyzed in Figure 4F and 4G. when lung cancer A549 cells with stable silencing Ang-2 transcription in the shRNA group, the level of the epithelial E-cadherin expression was significantly (P<0.001, Figure 4F) increasing than that in the NC or control group; otherwise the levels of mesenchymal VIM, and transcriptional factor Snail and Twist expressions were significantly (P<0.05, Figure 4G) decreasing than those in the NC or control group.

DISCUSSION
Lung cancer is the most common malignant tumors with and complex pathogenesis [2]. Its occurrence, development, and metastasis are associated with angiogenesis formation that make cancerous cells more invasive and affect chemotherapy and radiotherapy. Angiogenesis is a fundamental process involving a variety of pathological processes and sustains the progression of many neoplastic diseases [6]. Moreover, it may enhance cancer cell proliferation and anti-apoptosis, and facilitate metastasis [21][22][23]. Ang-2 upregulated in multiple inflammation-related tumors and associated closely with lung cancer progression. Although abnormal Ang-2 expression has been reported as useful biomarker for patients with lung cancer [24], however, the relationship between Ang-2 and tumor metastasis still needs to be identified. Therefore, the aims of this present study were to investigate the Ang-2 expression and its clinicopathological characteristics in tumor tissues, and silenced Ang-2 transcription by specific shRNA to analyze the metastasis mechanism and biological behaviors of lung cancer.
The instability structure of tumor vascular, extreme hypoxia and drug delivery problems, these changes promote tumor progression, invasion of surrounding tissue, and distant metastasis [25]. Tumor tissue hypoxia causes cancer cells and macrophages to produce a large number of angiogenic factors that induce angiogenesis for the growth and development of lung cancer [26]. The expression of Ang-2 was increased in the peritumoral vascular remodeling area when capillary without   formation, and might be involved in the initiation and regulation of lung cancer angiogenesis [27]. In this study, the level of Ang-2 expression in lung cancer tissues was significantly higher than that in their paracancerous tissue (Figure 1), and closely positive correlated with tumor size, TNM staging, poor differentiation, and 5-year survival rate ( Figure 2), but there was no obvious relationship with patients' age and lymph node metastasis (Tables 1&2), indicated that high Ang-2 expression could be an independent predictor related to distant metastasis for patients' poor prognosis [28,29]. When lung cancer cells proliferate, tumor volume and oxygen consumption increase significantly, the cancer cells become hypoxic and overexpress HIF-1α, which leads to the secretion of angiogenic factors (Ang-2, VEGF and so on) and induces angiogenesis [30,31]. HIF-1α is suggested to be an important upstream molecule mediating VEGF expression and angiogenesis [16,25]. However, it remains partially unclear that Ang-2 effects on the progression of lung cancer. In this study, the best inhibitory effect of shRNA sequence was screened through transfecting over-expression Ang-2 cell lines ( Figure 3). Ang-2 expression in lung cancer (SPC-A-1, NCI-1650, A549, and NCI-1975) cells at protein or mRNA level was significantly higher than that in Beas-2B cells, especially in A549 cells. Among the shRNAs transfected into A549 cells, the effective shRNA-1 with significantly downregulating Ang-2 and inhibiting cell proliferation was chosen to analyze the biological behaviors of lung cancer cells. Metastasis progress of lung cancer consists of distinct steps, including tumor angiogenesis, cell detachment, intravasation, micrometastasis formation, and growth [32][33][34]. After transfected with shRNA-1, the A549 cells showed lower proliferative abilities than those in the NC-or the control-group at a time-dependent manner ( Figure 4). The abilities of the A549 cell migration and invasion in the shRNA group were lower than those in the NC-or control-group. Down-regulating Ang-2 affected the invasion and migration of A549 cells (Table 3), the data demonstrated that silencing Ang-2 transcription could remarkably alter the biological behaviors with decreasing migration and invasion abilities of lung cancer cells.
EMT is a process that increases the metastasis and invasive potential of tumor cells, with losing epithelial cell characteristics and acquiring a mesenchymal phenotype, including decreasing epithelial markers (cytokeratin or E-cadherin) and up-regulating mesenchymal indicators (VIM or N-cadherin) or transcription factors (Twist, Snail) [35][36][37]. Usually, inhibiting E-cadherin can induce  . (A & B), the cell proliferation ability was examined using CCK-8 at 450 nm. (C), the cell patterns (×200 magnification) with the crystal violet staining solution: c1, the cells in the shRNA group; c2, the cells in the NC group, and c3, the cells in the control group; (D), the cell migration abilities were presented as total number of cells that migrated to the bottom chamber without or with the transwell-precoated matrigel, as calculated at 6 random fields; (E), the cellinvasion abilities were presented as total number of cells that migrated to the bottom chamber without or with the transwell-precoated matrigel, as calculated at 6 random fields; (F), the EMT-related epithelial indicators and mesenchymal markers after silencing Ang-2 transcritpion of A549 cells were analyzed by the Western blotting with β-actin as loading control; and (G), the relative ratio from each protein to β-actin (n=3) compared with the control group. Ang-2, angiopoietin-2; shRNA, the Ang-2-shRNA transfection group; NC, the negative control group; and Control, the blank control group; * P<0.05. ** P<0.001.
In conclusion, the abnormal Ang-2 expression promoted cell proliferation and affected on abilities of invasion, metastasis, and EMT of lung cancers that should provide a new mechanism insight into tumor metastasis by inhibiting E-cadherin and upregulating VIM, Twist and Snail signaling ( Figure 5). Ang-2 was specifically overexpressed in tissues of lung cancers but not in benign lung diseases, and silencing Ang-2 transcription by specific shRNA or anti-human Ang-2 antibodies could effectively inhibit tumor growth or metastasis with related EMT formation. However, there was a certain limitation that a relatively small sample of patients was involved in this study. Although the associations between Ang-2 and tumor metastasis were discovered, it was necessary to conduct more in vitro and in vivo assays to explore the exact function and mechanisms [41]. Further work should be explored combination of specific shRNA plus multi-targeting strategies for effective lung cancer therapy [42,43].

Lung tissues
For this study, fresh lung cancerous-and selfcontrolled paracancerous-tissues were obtained from 122 lung cancer patients who under-went surgery at the Affiliated Hospital of Nantong University, China, from Jul. 2009 to Dec. 2011. All cases of lung cancer were confirmed clinical and pathological evidence. All patients had not received radiotherapy or chemotherapy before operation. Lung specimen were collected from cases (59 men and 63 women) diagnosed with adenocarcinoma according to histological classification. Tumor node metastasis (TNM) staging of the patient cohort included 83 cases at stage-I, 17 with II, and 22 with III-IV. The tissue specimen were immediately frozen in liquid nitrogen and kept at −80°C until required. Each of the specimen was divided into two parts: one was used for immunohisto-chemical (IHC) staining and the other was fixed in formalin, embedded in paraffin, and then cut into 4-μm-thick section for pathological examination with hematoxylin & eosin (H&E) staining. This present study was approved by the Ethics Committee permission (TDFY2011008) of the Affiliated Hospital of Nantong University, and all patients signed informed consent. The related important clinical information of each patient was collected from their medical records.

Cell lines and cell culture
Human lung epithelial Beas-2B cell line, non-small cell lung cancer (NSCLC) NCI-H1650, SPC-A-1, A549, and NCI-H1975 cell lines were obtained from the Institute of Biochemistry and Cell Biology, Chinese Academy of Science, Shanghai, China. All cell lines were cultured in RMPI-1640 medium (Corning, USA) containing 10 % fetal bovine serum (FBS, BI, Israel) and maintained in a 5% CO 2 humidified atmosphere at 37°C. The expression and gene transcription of Ang-2 in human lung Beas-2B cells, lung cancer SPC-A-1, NCI-1650, A549, and NCI-1975 cell lines were analyzed at protein level by Western blotting and at mRNA level by quantitative real-time reverse transcription PCR (qRT-PCR).

Total RNA extraction and qRT-PCR
Total RNAs from different cell lines were extracted using Trizol reagent (Invitrogen, USA). Reverse transcription to cDNA was carried out by using a Revert AidTM First Strand cDNA synthesis kit (MBI Fermentas, Vilnius, Lithuania USA) according to manufacturer's instructions. In brief, in a 20 μL reaction mixture containing 2 μg RNA and oligo primers, at 42°C for 1 h, and synthesized cDNA was used for PCR by using SYBR®Premix Ex TaqTMII (TaKaRa, Dalian, China) with primers. The primers used for real-time RT-PCR purchased from Genechem Co., LUd (Nanjing, China) were as follows: Ang-2 forward, 5′-ACGGCTGTGATGATAGAAATAGG-3′ and reverse, 5′ GTAGTTGGATGATGTGCTTGTC-3′; glyceraldehyde-3-phosphate dehydrogenase (GAPDH): forward, 5′-CAAGGTCATCCATGACAACTTTG-3′; and reverse: 5′-GTC CACCACCCTGTTGCTGTAG-3′ (NM_017008). Cycling conditions were as follows: 2 min of preincubation at 95°C, 10s of denaturation at 95°C, 30s of annealing at 60°C, and 45s of extension at 72°C for 40 cycles with iCycler (Bio-Rad. USA). The fluorescent product was detected at the end of each cycle. Melting curves analysis was performed to determine PCR efficiency and purity of the amplified product after completed qRT-PCR, collecting fluorescence between 70°C and 95°C at 0.5°C increments. For data analysis, the raw threshold cycle (C t ) value was normalized to the GAPDH for each sample to get ΔC t . The normalized ΔC t was then calibrated to control cell samples to get ΔΔC t . The relative quantitative results of mRNA was calculated by the equation 2 −ΔΔCt (n=3).
In short, cells were seeded in 6-well plates at a density of 4 × 10 5 cells/well, so that they could become about 60% confluence the next day. Transfection of shRNA plasmids according to manufacturer's instructions, the one of the most effective silencing sequences were screened according to Ang-2 expression at protein level by Western blotting or at mRNA transcriptional level by qRT-PCR analysis.

Cell proliferation assays
For analysis of cell proliferation, lung cancer A549 cells with Ang-2 overexpression divided into three groups: Ang-2-shRNA1 (shRNA1), negative-shRNA control (NC), and blank control group (Control) with total 5×10 3 cells in 100μL of medium into 96-well plates (n=3/per group). Cell Counting Kit-8 (CCK-8, DOJINDO, Japan) was used to evaluate cell proliferation according to the manufacturer's instructions. In brief, 10μL of CCK8 solution was added to the culture medium in each well and incubated for 2h. The cell absorbance at 450nm after 0h, 24h, 48h and 72h was determined, and the assays were repeated three times with triplicate samples.

Transwell migration and invasion assays
For cell invasion assays, modified Boyden Chambers consisting of Transwell-precoated Matrigel membrane filter inserts with 8 μm pores were used in 24-well tissue culture plates (BD Biosciences, Bedford, MA). Cells (5 × 10 3 ) from the shRNA, NC, and control groups (n=3/per group) were plated onto the top of the chamber in 100 μL of RMPI1640 medium without FBS and the bottom chamber was filled with RMPI1640 medium containing 10% FBS as a chemoattractant. After 24h of incubation in a 5% CO 2 humidified chamber at 37°C, non-invading cells were removed by wiping the upper surface of the membrane with a cotton swab, and the filter membrane was fixed with 4% paraformaldehyde and stained with Crystal Violet Staining Solution. The degree of invasion was quantified by counting the cells that had migrated through the membrane in at least six random fields (total magnification, ×200) per filter. Experiments were repeated three times in triplicate. For analysis of cell migration, we use the modified Boyden Chambers without the Transwell-precoated Matrigel membrane filter with the method performed as above.

Immunohistochemistry (IHC)
TMA slides used for IHC analysis were deparaffinized, and eroxidase was quenched with methanol and 3% H 2 O 2 for 15 min. For antigen retrieval, the sections were boiled under pressure in citrate buffer (pH 6.0) for 3 min and then incubated for 120min with primary mouse anti-human Ang-2 (ab155106, Abcam, USA), washing with phosphate buffered saline (PBS), incubated with horse reddish peroxidase (HRP)conjugated goat anti-mouse IgG (A21020; Abbkine, USA) for 15 min at 1:5000 dilution, and washed again with PBS. In finally, the slides were incubated with diaminobenzidine and counterstained with hematoxylin solution, dehydrated in ethanol, cleared in xylene, and cover-slipped. For the negative control reactions, the primary antibody was instead with PBS. www.impactjournals.com/oncotarget

Evaluation of IHC findings
The results of IHC staining were assessed by two independent pathologists without knowledge of the clinicopathologic features. The percentages for Ang-2 positive cells were scored as follows: 0 (0%), 1 (1% ~ 33%), 2 (34% ~ 66%), and 3 (67% ~ 100%) according to the previous reported method [44]. Staining intensity was stratified into four categories: 0 (negative), 1 (weakly), 2 (moderate), and 3 (strongly). The sum of the percentage and intensity score was defined as the IHC staining score. According to above criterion, the lung tissues with Ang-2 expression were divided into two groups: low with 0 ~ 1 scores and high 2 ~ 3 scores. The higher score for positive percentages and staining intensity was taken as the final score when there was a difference between duplicate tissue scores.

Statistical analysis
Statistical analysis was carried out by using SPSS software (version 20.0). Results of cell proliferation, transwell migration and invasion were expressed as the mean ± standard deviation (SD). Difference between groups or clinicopathological factors was evaluated by the χ 2 or t test. Survival curves were made using the Kaplan-Meier method and analyzed by the log-rank test or the Cox regression. Significance level was set at P value < 0.05 for all tests.

CONFLICTS OF INTEREST
No potential conflicts of interest were disclosed.