Up-Regulation of Insulin-Like Growth Factor-2 Promotes Cell Proliferation and Migration and is Associated with Bone Metastasis in Non Small Cell Lung Cancer.

Background:Non-small cell lung cancer (NSCLC) is a malignant tumor with high morbidity and mortality. About 40% NSCLC patients develop bone metastasis. The aberrant expression of IGF2 (insulin-like growth factor-2) is known to be linked with many malignancies, but it has not been reported in NSCLC. Methods and Results:We rst used the GEO database for differential gene analysis and showed a correlation between IGF2 and bone metastasis in NSCLC. IGF2 was highly expressed in primary NSCLC and even more highly expressed in bone metastases. Our in vivo experiment also showed that IGF2 promoted lung cancer proliferation and migration. In addition, IGF2 overexpression activated AKT/NF-κB pathway by promoting IGF1R and INSR phosphorylation, and enhanced osteolytic lesions in lung cancer. Conclusions(cid:0)Up-regulation of insulin-like growth factor-2 can promote the proliferation and migration of non-small cell lung cancer cells, and is related to bone metastasis.


Introduction
Lung cancer is one of the most frequently diagnosed malignancies and the primary cause of cancerassociated mortality [1]. Non-small cell lung cancer (NSCLC) including adenocarcinoma, large-cell carcinoma along with squamous cell carcinoma accounts for more than 85% of all lung cancer cases [2].
Bone metastasis occurs in 30-40% of individuals with metastatic lung cancer, causing pathological fractures, hypercalcemia, severe pain, as well as nerve compression [3]. The mean survival of individuals with lung cancer after diagnosis with bone metastasis is often less than 6 months. Thus, limiting bone metastasis is a crux in the clinical treatment of NSCLC [4].
Insulin-like growth factor 2 (IGF2), a polypeptide growth factor family member, is expressed in most tissues [5].Numerous reports suggest that IGF2 modulates cell growth, differentiation, as well as metabolism. There is a close relationship between IGF2 and the development of cancers [6]. Previous researches have documented that IGF2 is abnormally expressed in various human malignancies, and high expression of IGF2 is usually linked to a poor clinical outcome [7][8][9][10]. Nevertheless, the role that IGF2 plays in lung cancer and the subsequent cascade of IGF2 remain unclear.
Herein, we illustrated that IGF2 expression was sequentially increased in non-malignant lung tissues, primary NSCLC, as well as NSCLC bone metastasis. Further examination showed that overexpression or knockdown of IGF2 could markedly promote or reduce proliferation, migration along with in ltration of NSCLC

Cell transfection
Cells in the log growth phase were used once they reached 5070% con uence. The DNA Transfection Reagent (Bimake) was used according to the operation manual 2.5 Immunohistochemistry (IHC) Tissues were xed in 4% PFA at 4˚C for 24 hours, and sectioned into 4-mm slices, followed by embedding in para n. Afterwards, 2% hydrogen peroxide dispersed in methanol was employed to block the endogenous peroxidase activity. Thereafter, the slides were inoculated in citrate buffer (pH 6.0) and then heated in a microwave oven for 12 minutes to retrieve the antigen. Subsequently, the slides were inoculated with rabbit anti-human IGF2 antibody (cat. no. ab9574; Abcam, Cambridge, MA, USA; 1:1,000).
Thereafter, the samples were inoculated with the goat anti-rabbit secondary antibody (Cat No, Dako, Glostrup, Denmark; 1:2000). Then, DAB staining (Cat No, Dako) was done as described by the manufacturer. Finally, light counterstaining of the samples with hematoxylin was done, followed by mounting. Two experienced pathologists assessed the results and images were acquired.

Colony formation assay
Cells were digested, counted, planted in 6-well plates in triplicate, grown for 14 days, washed with PBS, xed in 4 % PFA, followed by 0.1 % crystal violet staining (Sigma, St Louis). The number of effective colonies consisting of more than 50 cells was counted.

Cell cycle assay
Five hundred A549 or PC9 cells were inoculated in 6-well plates and inserted with pcDNA3.1+, or OE-IGF2 plasmids via transfection 48 hours. We harvested the cells, followed by xing them with ice-cold 70% ethanol. Thereafter, the cells were rinsed in PBS, and rehydrated in PBS enriched with propidium iodide (0.5 mg/mL dispersed in PBS with 0.1% sodium azide) and RNase A (1 mg/mL) for 30 min in the dark. Finally the FACSCalibur ow cytometer (Becton-Dickinson, USA) was employed to analyze the samples.
Lastly, the FlowJo software (Tree Star, USA) was employed to analyze the data.

Transwell migration and invasion assay
Page 5/20 The 8-µm pore size transwell insert without matrigel (3422, Corning, USA) and with matrigel (354480, Corning, USA) were utilized for transwell migration and in ltration assay respectively. Digestion of the cells was done, followed by counting. Overall, 1 × 10 5 cells in 100 µL medium enriched with no FBS were inoculated in the upper inserts and 500 µL medium augmented with 10% FBS was inoculated in the lower inserts as chemoattractant and incubated for 24 hours. After that, we removed the cells in the upper inserts, while the bottom surface-cells were xed with 4% PFA, followed staining with 0.1% crystal violet for 15 minutes. After that, 5 random elds were counted with a microscope at 400 ×. Lastly, acetic acid elution was performed and the OD values read at 570 nm.

Western blot
The RIPA buffer (KeyGen Biotech, Shanghai, China) enriched with PMSF (KeyGen Biotech), protease inhibitor, as well as phosphatase inhibitor cocktail, was employed to lyse the cells. The Western blotting was conducted as documented previously [20]. 20

Statistical analysis
Results are given as means ± SEM unless speci ed. Two-tailed unpaired t-test was employed to compute signi cant differences between the mean values of groups. The Kaplan Meier plotter website (http://kmplot.com/) was used for OS (overall survival) and PFS (progression-free survival) analysis in NSCLC patients. The IGF2 probe set was 202409_at (although there were three IGF2 probe sets, the results were similar.) The patients were split on the basis of the median, and the univariate cox regression was performed. P < 0.05 signi ed statistical signi cance.

Results
3.1 IGF2 is upregulated in NSCLC bone metastasis and negatively correlated with clinical prognosis.
Herein, gene expression patterns from two datasets in lung adenocarcinoma were selected for comparison of gene expression. According to the introduction, the samples were divided into three groups. After that, an overlapping analysis of the differential expression genes (DEGs) between the groups was performed. DEGs were de ned as genes with P < 0.05 along with the absolute fold change > 1.5. The top 30 DEGs are shown in Fig. 1A. It was found that the expression level of IGF2 in the nonmalignant lung tissue was low but sequentially increased in the NSCLC and NSCLC bone metastasis tissues. Then, we used qRT-PCR to validate the mRNA expression of the top 10 DEGs in our own clinical specimens (5 non-malignant lung tissues, 5 primary NSCLC tissues and 5 NSCLC bone metastasis tissues). The mRNA expression level of IGF2 was very consistent with the microarray analysis (Fig. 1B). Besides, western blotting along with IHC analysis were employed to compare the expression of IGF2 in the above-mentioned specimens and revealed a similar change in IGF2 mRNA level (Fig. 1C-D). To assess the relationship of IGF2 expression with survival of individuals with the NSCLC, we explored IGF2 expression contribution to OS, as well as FPS of the patients in a clinical microarray data resource containing 1,648 adenocarcinoma patients. The result revealed that elevated expression of IGF2 predicted a poorer OS in individuals with adenocarcinoma (HR = 1.44, P = 0.000) (Fig. 1C-D). Moreover, patients harboring elevated IGF2 expression also exhibited a shorter PFS as in contrast with those with low IGF2 content (HR = 1.35, P = 0.000, (Fig. 1C-D)) 3.2 IGF2 overexpression promotes proliferation and colony formation in NSCLC cells, while IGF2 inhibition has an opposite effect.
To explore further the function of IGF2 in NSCLC, we subsequently up-and down-regulated IGF2 expression in NSCLC cells by transfecting overexpression or shRNA plasmid of IGF2. The suppression and overexpression of IGF2 were validated at mRNA, as well as protein levels. As illustrated in Fig. 2, the number of colonies was increased in A549, as well as PC9 cells transfected with OE-IGF2 and decreased in cells transfected with sh-IGF2. Additionally, we also performed cell cycle assessment by ow cytometry and found that IGF2 overexpression increased the fraction of cells in S and G2-M-phase, and IGF2 inhibition produced an opposite effect. These data suggest that IGF2 enhanced cell proliferation of NSCLC cells (Fig. 3).

IGF2 increases migration and in ltration of NSCLC cells in vitro.
The relationship of high IGF2 expression in bone metastasis with poor prognosis of NSCLC patients prompted us to investigate whether manipulation of IGF2 could alter the migration along with the in ltration potential of lung cancer cells. Therefore, we knocked down or overexpressed IGF2 in both A-549 and PC9 cells, and found that the migration and in ltration of A549 and PC9 cells were either enhanced or suppressed by IGF2 overexpression or knockdown (Fig. 4).

IGF2 overexpression promotes NSCLC cell-induced osteoclast differentiation.
As a kind of secreted protein, IGF2 could be secreted into the microenvironment of tumor cells. To explore further the role of IGF2 in NSCLC bone metastasis, we selected Raw246.7 cells to test the NSCLC cellinduced osteoclast differentiation after altering IGF2 expression. Conditional media from A549 cells inserted with OE-IGF2, sh-IGF2 and their corresponding Ctrl plasmids were employed as different stimuli during osteoclastogenesis. We evaluated the expression of IGF2 in the conditional media of A549 by ELISA. As shown in Fig. 5E, the IGF2 content in the conditional medium from OE-IGF2 A549 cells was signi cantly increased, while the conditional medium from sh-IGF2 A549 cells was markedly reduced. The TRAP staining assay indicated that the level of TRAP positive multinucleated osteoclast formation was accelerated in Raw246.7 cells inoculated with the conditional medium from OE-IGF2 A549 cells, while it was decreased in Raw246.7 cells inoculated with the conditional media from sh-IGF2 A549 cells (Fig. 5). We also detected the mRNA levels of markers of osteoclasts including TRAP, nuclear factor of activated T cells, NFATC1 (cytoplasmic 1), and CTSK (cathepsin K), and found that the mRNA content of NFATc1, TRAP as well as CTSK were consistent with the results of the osteoclast differentiation test.

IGF2 activates AKT/NFκB signaling in Raw246.7 cells by targeting IGFR1 and INSR
The AKT/NFκB signaling cascade plays an indispensable role in regulating osteoclast differentiation progression in Raw246.7 cells. Multiple detection assays showed that protein contents of NFATc1, TRAP and CTSK were markedly increased in Raw246.7 cells inoculated with IGF2 ( Fig. 6A-B). In addition, IGF2 addition remarkably increased the expression levels of phosphorylated IGFR1, INSR, AKT and NFκB in a dose-dependent approach. Knowing that IGFR1 and INSR are IGF2 receptors, we next evaluated whether knockdown of IGFR1 or INSR by siRNA could inhibit the AKT/NFκB signaling pathway. The result showed that both phosphorylation of AKT and NFκB and the protein level of NFATC1, TRAP and CTSK were partially suppressed after knockdown of IGFR1 or INSR respectively. The IGF2-induced AKT/NFκB phosphorylation and increase of osteoclast markers were markedly reversed by knockdown of IGFR1 and INSR by at the same time (Fig. 6C-E).

Discussion
Expression of the imprinted gene IGF2 is usually upregulated numerous human cancers, but the molecular mechanisms remain elusive [11][12][13][14]. Our current study rstly addressed the upregulation of IGF2 in NSCLC bone metastasis in contrast with primary tumors, and in addition identi ed the negative relationship of high expression level of IGF2 and poor prognosis in NSCLC patients. Our data substantiate that IGF2 overexpression signi cantly enhanced NSCLC cell growth, migration and in ltration, which is consistent with the results in other recent studies [15,16]. We also demonstrated that IGF2 triggered the ATK/NF-κB pathway through IGFR1 and INSR, thus enhancing osteoclastogenesis.
IGF2, a 7.5 kDa mitogenic peptide hormone, is secreted by the liver and numerous other organ tissues [5]. IGF2 is upregulated in many cancers and is linked to resistance to chemotherapy and dismal prognosis, [6] as represented by more rapid progression of the disease in chronic myeloid leukemia, shorter time to disease relapse in esophageal cancer and higher mortality in breast cancer, ovarian cancer and colorectal cancer [11,[17][18][19][20]. Recent research ndings also suggest that IGF2 expression facilitates lung tumorigenesis and indicates poor prognosis in NSCLC [10]. Jang et al demonstrated that psychologic stress activated the IGF-1R pathway by inducing exocytosis of IGF2 in lung epithelial cells and promoted lung tumorigenesis in mice [21]. These results were predictable, cause many studies have shown that as a mitogenic peptide hormone, secretes the function of IGF2 is regulating cell growth, differentiation and metabolism [5,10,22].
Bone is one of the most frequent organs of NSCLC metastasis. The occurrence of bone metastasis entails osteolysis and cancer cell growths [15,23]. To the best of our knowledge, the role of IGF2 in NSCLC bone metastasis remains unclear. Fukuoka et al documented that hypoxic stress promoted differentiation of osteoclasts through accelerating IGF2 generation by non-osteoclastic cells [24]. Nakao et al pointed out that stimulation of exogenous IGF2 from the initial stage was more pivotal relative to the later stage for the promotion of osteoclastogenesis [25]. Our results indicate that NSCLC cells produced IGF2 in an autocrine or paracrine manner, auutocrine secretion of IGF2 stimulated cancer cell proliferation and migration, and paracrine secretion of IGF2 enhanced osteoclast differentiation of Raw 246.7 cells. Murayama [15] .Given the potential impacts of IGF2 in the bone microenvironment and progress of cancer, we suggest that IGF2 participates in the onset of bone metastasis and could be a prospective prognostic biomarker and targeting opportunity for NSCLC.
Similarly to IGF1, IGF2 is known as a modulator of G1 to S phase transition [27]. IGF2 promotes cancer cell growth and epithelial-mesenchymal transition (EMT) through mainly IGF1R [28]. IGF1R is a receptor tyrosine kinase and signals downstream to the PI3K/AKT and mTOR pathways [6]. There are other types of IGF2 receptors. Indeed, we found that INSR, a G protein coupled receptor, is also a candidate receptor of IGF2, which mediates the activation of the PI3K/AKT casacde. The classic NF-κB signaling cascade is known to play an indispensable role in osteoclast differentiation [29]. Ethics approval and consent to participate The present study was approved by the Ethics Committee of our center and written informed consent was obtained from the surviving patients, or family members of those who had succumbed.
Patient consent for publication Not applicable.