IL‐6 promotes metastasis of non‐small‐cell lung cancer by up‐regulating TIM‐4 via NF‐κB

Abstract Objectives Interleukin‐6 (IL‐6) is critical for the development of non‐small‐cell lung cancer (NSCLC). Recently, we identified T‐cell immunoglobulin domain and mucin domain 4 (TIM‐4) as a new pro‐growth player in NSCLC progression. However, the role of TIM‐4 in IL‐6‐promoted NSCLC migration, invasion and epithelial‐to‐mesenchymal transition (EMT) remains unclear. Materials and Methods Expressions of TIM‐4 and IL‐6 were both evaluated by immunohistochemical staining in NSCLC tissues. Real‐time quantitative PCR (qPCR), Western blot, flow cytometry and RT‐PCR were performed to detect TIM‐4 expression in NSCLC cells with IL‐6 stimulation. The roles of TIM‐4 in IL‐6 promoting migration and invasion of NSCLC were detected by transwell assay. EMT‐related markers were analysed by qPCR and Western blot in vitro, and metastasis was evaluated in BALB/c nude mice using lung cancer metastasis mouse model in vivo. Results High IL‐6 expression was identified as an independent predictive factor for TIM‐4 expression in NSCLC tissues. NSCLC patients with TIM‐4 and IL‐6 double high expression showed the worst prognosis. IL‐6 promoted TIM‐4 expression in NSCLC cells depending on NF‐κB signal pathway. Both TIM‐4 and IL‐6 promoted migration, invasion and EMT of NSCLC cells. Interestingly, TIM‐4 knockdown reversed the role of IL‐6 in NSCLC and IL‐6 promoted metastasis of NSCLC by up‐regulating TIM‐4 via NF‐κB. Conclusions TIM‐4 involves in IL‐6 promoted migration, invasion and EMT of NSCLC.


Non-small cell lung cancer (NSCLC) is one of the malignant tumours
with the fastest increase in mortality despite the implementation of novel targeted therapies and chemotherapeutic regimes. 1,2 Metastasis is responsible for the poor prognosis of NSCLC. 3,4 Epithelial-mesenchymal transition (EMT), as a key regulator of metastasis, involves a series of phenotypic and behavioural changes, which contributes to the transformation of early tumours into invasive malignant tumours. 5,6 EMT has been verified to be responsible for the development of NSCLC resistance to anti-cancer agents. [7][8][9] Therefore, early identification of novel prognostic molecular markers related to NSCLC metastasis and EMT is urgently required.
Our previous studies showed that T-cell immunoglobulin domain and mucin domain 4 (TIM-4), closely related to poor NSCLC prognosis, could promote the growth, proliferation and cell cycle progress of NSCLC cell lines through the interaction of its RGD motif. 10 However, the role of TIM-4 in NSCLC metastasis and EMT has not been reported. Moreover, TIM-4 levels in NSCLC tissues are significantly elevated, while the expression levels in various lung cancer cell lines are relatively low, 10 and the relevant mechanism is still unclear.
Interleukin-6 (IL-6) is an important cytokine for EMT and tumour metastasis of NSCLC. [11][12][13] The increase of IL-6 can cause the resistance to molecular targeted therapy in lung cancer, [14][15][16] and circulating IL-6 level may be a prognostic marker for patients with NSCLC. 17 Moreover, a humanized anti-IL-6 antibody (ALD518) has been used to treat NSCLC. 18 Our previous studies showed that IL-6 could induce TIM-4 expression in vitro. 10 Here, we investigated the exact role of TIM-4 in IL-6-induced NSCLC migration, invasion and EMT.

| Animals
Pathogen-free BALB/c nude mice (5-6 weeks old, male) bought from Vital River Laboratory Animal Technology Co., Ltd were randomly assigned to experimental groups and housed in home cages with an alternating 12 hours light and 12 hours dark cycle, water available, in the Animal Center of Shandong University under specific pathogen-free conditions. All animal experiments were performed with the approval of the Committee on the Ethics and use of Animal Experiments of Shandong University.

| Human samples
Non-small-cell lung cancer tissues samples were collected from 106 patients who had undergone surgery at Zhoushan Hospital (Zhejiang, China) from January 1 in 2011 to January 1 in 2014, without receiving preoperative chemo-or radiotherapy. The patients' age, gender, smoking history, carcinoembryonic antigen (CEA), tumour size, histology type, pleural invasion, lymph node metastasis, stage, and grade were determined by a reviewer according to the medical records. Each tumour sample was classified according to the 8th edition tumour-nodemetastasis (TNM) classification of lung cancer. Follow-up duration was determined from the date of surgical treatment until the death. The study was approved by the ethics committee of Zhoushan Hospital.

| In vivo lung cancer metastasis test
Migratory ability was determined with lung cancer metastasis test in BALB/c nude mice. Briefly, A549 cells were transfected with LV-NC-Luciferase (LV-NC) or LV-shTIM-4-Luciferase (LV-shTIM-4), respectively. Stable cell lines with TIM-4 knockdown or control were obtained by a high concentration of puromycin selection (12 μg/ mL) for 2 weeks, and subsequently, stable clones were maintained in culture with a low concentration of puromycin (1 μg/mL) for tumour bearing in BALB/c nude mice. Then, a tail-vein injection model administration intraperitoneally (n = 3). And the dose of IL-6 (dissolved in PBS) was 2.5 μg/kg once daily at 10 o'clock in the morning for the first 2 weeks, then 5 μg/kg once every 3 days throughout the next 4 weeks. After 6 weeks, mice were monitored using the IVIS Spectrum In Vivo Imaging System (PerkinElmer) of Advanced Medical Research Institute, Shandong University. Mice were intraperitoneally injected with 200 μL of D-luciferin (150 mg/kg body weight, PerkinElmer) and anesthetized with isoflurane gas after 5 minutes. One minute later, mice entered the state of general anaesthesia, and then, they were transferred into the cage of the in vivo imaging system. The chest and abdomen of mice were completely exposed, and then, image calibration and visualization were performed using Living Image 4.2 software (PerkinElmer). At 7 weeks, mice were euthanized with an intraperitoneal injection of pentobarbital-phenytoin solution in the lower right abdominal quadrant. Then lungs were isolated, weighed, fixed in 4% paraformaldehyde at least for 24 hours and embedded in paraffin. Pathological changes were evaluated in sections (5 μm) by H&E staining. Numbers of metastatic nodules per lung were counted by panoslice scanner system (Tykor).

| Statistical analysis
GraphPad Prism 6.0 software was used for data analysis. Log-rank test was utilized to compare the survival for different groups. A logistic regression model was applied to investigate the independent predictive factors responsible for TIM-4 expression. Additionally, the correlation between TIM-4 and IL-6 was analysed by Spearman's rank correlation. P-values of <0.05 were considered as statistically significant. Other quantitative values were presented as the mean ± SD. Represent data were from at least triple or duplicate independent experiments. The differences in mean values between two groups were analysed by the Student's t test. *P < .05, **P < .01, ***P < .001, ****P < .0001, ns: not significant.
Additional supporting information of MATERIALS AND METHODS could be found in the Appendix S1.

| High expression of TIM-4 was positively correlated with IL-6 in NSCLC, which indicated poor prognosis
Previously, we found that TIM-4 was highly expressed in cancer tissues of NSCLC. 10 Here, we further investigated the predictive factors for TIM-4 expression in clinical samples of NSCLC. Multivariate logistic regression analysis showed that tumour size (>3 cm) (OR, 2.574; 95% CI, 1.646-4.026; P = .032), lymph node metastasis (OR, 2.011; 95% CI, 1.472-3.744; P = .027) and high IL-6 expression (OR, 2.951; 95% CI, 1.082-6.085; P = .038) were independent predictive factors for TIM-4 expression (Table 1). Subsequently, the correlation between TIM-4 and IL-6 were analysed separately, and it was shown that TIM-4 expression was positively correlated with IL-6 expression in NSCLC ( Table 2, r = .489, P = .021). Representative stained fields on histopathological slides for TIM-4 and IL-6 were displayed in Figure 1A. Additionally, the combination of TIM-4 and/or IL-6 expressions was used to explore the relationships with prognosis. As shown in Figure 1B, patients with high TIM-4 and IL-6 expression showed the worst prognosis, while patients with TIM-4 and IL-6 double low appeared the best prognosis (P = .0282). These results suggested that TIM-4 might be involved in the process of IL-6 promoting the development of NSCLC.  Figure   S1A). Subsequently, 50 ng/mL IL-6 was used to stimulate lung cancer cells for 24 hours. Above all, the results showed that TIM-4 expression in lung cancer cell lines was up-regulated after IL-6 stimulation.

| IL-6 promoted TIM-4 expression in NSCLC cell lines via NF-κB pathway
It was reported that TIM-4 inhibited cytokine production via NF-κB signalling pathway 19 and had no effect on STAT3 phosphorylation, 20 while IL-6 could increase the activation of NF-κB 16 and STAT3 signalling pathway. 21 We then tested the changes of these signal molecules in IL-6-induced up-regulation of TIM-4 in lung cancer cells with NF-κB inhibitor or STAT3 inhibitor, respectively. The results revealed that IL-6 could increase the phosphorylation of p65  F I G U R E 2 IL-6 promoted TIM-4 expression via NF-κB pathway. IL-6 was used to stimulate A549 and H1975 cells. TIM-4 mRNA and protein levels were detected by qPCR (A), Western blot (B) and flow cytometry (C), respectively. D, NF-κB or STAT3 inhibitor was used to incubate with IL-6 stimulated A549 or H1975 cells, and phosphorylation of P65 or STAT3 and TIM-4 protein expression were detected by Western blot. E, In A549 and H1975 cells, the TIM-4 promoter activity was measured using a dual fluorescent reporter assay after stimulation with IL-6, and IL-6 plus NF-κB inhibitor, respectively. The box plots in A, C and E showed median ± SD of three independent experiments. ns: no significance, *P < .05, **P < .01, ***P < .001, ****P < .0001, by 2-tailed Student's t test analysed the transcriptional factors associated with NF-κB components and binding sites in TIM-4 promoter (−1247 to +300 bp) by PROMO software and JASPAR software ( Figure S1B). In accordance with the above prediction results, the effect of IL-6 on promoting TIM-4 promoter activity was attenuated after the addition of a specific inhibitor of NF-κB ( Figure 2E).

| TIM-4 increased IL-6 production
To detect whether TIM-4 knockdown affects the expression of IL-6, we detected the attenuated IL-6 expression in transcriptional level by RT-PCR after interfering TIM-4 (LV-shTIM-4-GFP) in A549 cells ( Figure S4A). We then tested the changes of signal molecules NF-κB and STAT3 in A549 cells overexpressed TIM-4 with NF-κB inhibitor or STAT3 inhibitor, respectively. The results revealed that TIM-4 promoted the phosphorylation of STAT3 and secretion of IL-6 in A549 cells, but not phosphorylation of P65 ( Figure S4B).
However, TIM-4-induced up-regulation of IL-6 in A549 cells was slightly decreased in STAT3 inhibitor treatment group, but no change was found in NF-κB inhibitor group ( Figure S4B). Above all, TIM-4 could also increase IL-6 production in lung cancer cells, which might form a positive feedback loop in IL-6 involved lung cancers.

| IL-6 promoted metastasis of NSCLC by upregulating TIM-4 via NF-κB in vitro
Consistent with the report, 22 we found that IL-6 promoted migration and EMT of lung cancer cells ( Figure S5A

| TIM-4 knockdown inhibited IL-6-enhancing migration and invasion in lung cancer cells in vivo
The lung metastasis mice model was established to investigate  as PD-L1, which further aggravated the tumour development. [28][29][30] Therefore, it was worth exploring whether TIM-4 expression was regulated by IL-6 in NSCLC cells. In clinical samples with NSCLC, we found that IL-6 was an independent predictor of TIM-4 expression, and patients with TIM-4 and IL-6 double high expression showed the worst prognosis. Importantly, we found that IL-6 indeed increased TIM-4 expression in lung cancer cell lines.

| D ISCUSS I ON
Next, we further examined the mechanism by which IL-6 induced the expression of TIM-4. As we know, NF-κB is considered to be a central mediator of immune and inflammatory response, and IL-6 could increase the activation of NF-κB signalling pathway. 31 bound to integrin to promote lung cancer growth. 10 As expected, our study found that IL-6 promoted TIM-4 expression in NSCLC cell lines via NF-κB pathway. In addition to NF-κB, STAT3 is another key factor in IL-6 signalling pathway, 34,35 and it has been documented that IL-6 promoted lung cancer development and progression by activating STAT3. 36,37 Another study found that IL-6 could induce immunosuppressive factor PD-L1 expression on peripheral myeloid cells through STAT3-dependent mechanism. 38  showed that TIM-3 facilitated osteosarcoma proliferation and metastasis through the NF-κB pathway and EMT. 40   University, for their support.

CO N FLI C T O F I NTE R E S T
The authors declare no conflict interest. L. and X. Y. performed the other experiments and analysed the data.

AUTH O R CO NTR I B UTI
X. L. and C. M. helped to design the experiments.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available within the article and its Appendix S1 or on reasonable request from the corresponding author.