IGF-1 induces the epithelial-mesenchymal transition via Stat5 in hepatocellular carcinoma

It has been reported that the epithelial-mesenchymal transition (EMT) plays an important role in hepatocellular carcinoma (HCC). However, the relationship between the insulin-like growth factor-1 (IGF-1) and EMT of HCC was not fully elucidated. In the present work, we found that the expression of N-cadherin, Vimentin, Snail1, Snail2, and Twist1 was positively associated with IGF-1R expression, while E-cadherin expression was negatively associated with IGF-1 expression in human HCC samples. Furthermore, we observed that IGF-1 up-regulated the expression of N-cadherin, Vimentin, Snail1, Snail2 and Twist1, and down-regulated the expression of E-cadherin. In addition, Stat5 was induced in IGF-1-treated HepG2 and Hep3B cells, and Stat5 inhibition or siRNA significantly affected IGF-1-induced EMT in HepG2 and Hep3B cells. In conclusion, IGF-1 induces EMT of HCC via Stat5 signaling pathway. Thus, IGF-1/Stat5 can be recommended as a potential and novel therapeutic strategy for


INTRODUCTION
Hepatocellular carcinoma is the leading factor in cancer-associated deaths all over the world [1][2][3]. Therapy outcome and prognosis for patients with HCC were demonstrated to be very poor in spite of operation combined with radiotherapy. In recent years, the 5-year overall survival rate of HCC patients accounts for nearly 15% [4]. In addition, patients suffered from different situations, iincluding resistance to chemo-radiotherapy, metastatic tendencies, and diagnosis at advanced stages, which lead to low survival rate and poor life status [5][6][7][8]. Thus, it is essential to hunt for a novel way to minimize the progression of HCC.
IGF family and their receptors play an important role in the development of tumors [9]. Recently, the role of IGF-1/IGF-1R signaling in human solid tumors has been widely reported, including lung cancer, gastric cancer [10,11]. Recent reports have indicated that IGF-1 facilitates the progression of lung cancer by regulating cell proliferation, angiogenesis, immune evasion, and metastasis [12]. In addition, it should be noted that Stat5 was associated with IGF-1-induced angiogenesis in many types of cancers. Kalita A et al demonstrated that IGF-1 induced upregulation of cyclin D1 expression in neural cell lines via the JAK/STAT5 signaling pathway [13]. However, no studies have been reported on the role of Stat5 in IGF-1-modulated EMT of HCC.
In this study, our study aims to find the relationship between IGF-1R and EMT biomarkers (E-cadherin, N-cadherin, Vimentin, Snail1, Snail2, Twist1) in HCC samples, and then we determined the impact of IGF-1 on cell proliferation, migration and invasion of HCC cells using in vitro assay. Finally, we knockdown the expression

Expressions and correlations of IGF-1R and EMT biomarkers in human HCC tissues
It has been reported that both IGF1R and EMT were involved in the metastasis of HCC. Firstly, we investigated the relationship between the expression IGF1R and EMT biomarkers in human HCC tissues using qRT-PCR assay. Our data demonstrated that the IGF1R expression was significantly negatively associated with E-cadherin expression in HCC tissues (p=0.014) (Figure 1a). In addition, the expression of IGF-1R was significantly positively associated with the expression of N-cadherin, Vimentin, Snail1, Snail2, and Twist1 (all p<0.05) (Figure 1b-1f), which indicates that IGF-1/IGF-1R axis seems to be involved in EMT of human HCC.

Stat5 expression is induced in IGF-1-treated HepG2 and Hep3B cells
Next, we conducted western blotting to explore whether Stat5 expression is induced in IGF-1-treated HepG2 and Hep3B cells. Our results suggested that IGF-1 induced an obvious upregulation of phosphorylated Stat5 in a time and concentration-dependent manner in HepG2 cells ( Figure 2a). Furthermore, IGF-1 induced an obvious upregulation of phosphorylated Stat5 in a time and concentration-dependent manner in Hep3B cells ( Figure  2b). At the same time, we determined 100 ng/ml of IGF-1 for subsequent assays.

Stat5 is involved for IGF-1-induced wound closure of HepG2 and Hep3B cells
To investigate the role and mechanism of IGF-1 in the migration of HepG2 and Hep3B cells, we conducted the wound healing assays. IGF-1 obviously promoted HepG2 and Hep3B cell wound closure in a time-dependent manner ( Figure 3). Moreover, it should be noted that the Stat5 inhibitor markedly prevented IGF-1-induced wound closure of HepG2 and Hep3B cells ( Figure 3).

Stat5 is involved in IGF-1-induced migration of HepG2 and Hep3B cells
We further confirmed the metastatic properties of IGF-1-treated HepG2 and Hep3B cells using migration assays. We found that HepG2 and Hep3B cell lines exhibited stronger migratory capacity in response to IGF-1 treatment than control ( Figure 4). Furthermore, the Stat5 inhibitor markedly prevented IGF-1-mediated migration of HepG2 and Hep3B cells ( Figure 4).

IGF-1 induces EMT via Stat5 in HepG2 and Hep3B cells
To further investigate whether IGF-1 promotes EMT progression in HepG2 and Hep3B cells, we utilized qRT-PCR and western blotting assays to evaluate the expression of EMT biomarkers after cells were treated with or without IGF-1. Our results demonstrated that the expression of E-cadherin mRNA was down-regulated, while the mRNA expression of N-cadherin, Vimentin, Snail1, Snail2 and Twist1 were up-regulated in IGF-1treated HepG2 and Hep3B cells compared with control ( Figure 5a, 5b). Furthermore, these findings above were validated by western blotting assay (Figure 5c), indicating that IGF-1 promoted EMT of HepG2 and Hep3B cells.
Next, in order to evaluate whether Stat5 knockdown is involved into IGF-1-mediated EMT of HCC, we treated cells with a Stat5 inhibitor or Stat5 siRNA before IGF-1 stimulation. And then EMT biomarkers were analyzed by qRT-PCR assay and western blotting assays. We found the Stat5 inhibitor obviously repressed IGF-1-mediated the mRNA and protein of EMT biomarkers in HepG2 and Hep3B cells (Figure 5a, 5b). In addition, qRT-PCR and western blotting analysis demonstrated that Stat5 siRNA effectively repressed the expression of Stat5 after posttransfection 48 h (Figure 6a, 6b). We found the Stat5 siRNA also repressed IGF-1-mediated the mRNA and protein of EMT biomarkers in HepG2 and Hep3B cells (Figure 7a, 7b). All in all, these results suggested that IGF-1 induces EMT via Stat5 in HepG2 and Hep3B cells.

DISCUSSION
IGF signaling pathway modulates energy metabolism and cell growth in normal conditions. In recent years, IGF-1R and its downstream signaling activation are demonstrated to be involved in the cancer cell proliferation, anti-apoptosis and metastasis in different cancers. It has been reported that the activation of and IGF-1R downstream signaling accelerate cancer progression through an enhancement of cell proliferation, migration and invasion; however, the role of IGF-1 in EMT of HCC was not reported. Our previous studies in cancer tissues led us to further explore the effects and corresponding mechanisms of IGF-1 on EMT profiles in the development of HCC.
It has been reported that IGF-1 inhibition repressed malignant cancer cell proliferation, migration and invasion in immunodeficient mice. IGF-1 inhibition also significantly inhibited the progression of EMT, with low expression of ZEB1 and N-cadherin and high expression of E-cadherin and MITF [14]. Yao C et al demonstrated the critical roles of IGF/STAT3/NANOG/Slug signaling www.impactjournals.com/oncotarget       pathways in the progression of CRC by regulating EMT and CSCs properties [15]. Li H et al also indicates that IGF-I induced EMT by the activation of a PI3K/Akt-GSK-3β-ZEB2 signaling pathway in the development of gastric cancer [16]. In addition, emerging reports suggests that EMT progression is directly modulated by EMT transcription factors, including Snail family, Twist1, or indirectly modulated by the Stat family pathway [17][18][19]. Notably, we found that Stat5 was induced by IGF-1 and that Stat5 knockdown obviously inhibited IGF-1induced HepG2 and Hep3B wound healing and migration. Therefore, our study may provide more useful information regarding the mechanism of HCC metastasis.
Based on molecular level, we further investigate whether Stat5 is also involved in IGF-1-induced EMT and found that Stat5 knockdown notably suppressed the IGF-1-induced increases in N-cadherin, Vimentin and EMT TFs expression (Snail1, Snail2, Twist1) and reversed IGF-1-mediated decrease in E-cadherin expression in HepG2 and Hep3B cells. Compared with other previous studies [20], Graham TR et al reported that IGF-I up-regulates the expression of ZEB1 in prostate cancer cells that exhibits a mesenchymal phenotype, however, ZEB1 inhibition abrogated the inhibition of E-cadherin protein, and further down-regulates the expression of N-cadherin and fibronectin proteins. As expected, ZEB1 inhibition also obviously decreases migratory and invasive potentials of prostate cancer cells than control.
In conclusion, IGF-1 signaling pathway induces EMT via Stat5 activation in HCC development. Thus, targeting of IGF-1/Stat5 or EMT may represent a novel way to decrease the development of HCC. Our work provides potential treatment strategy for HCC patients.

Ethics statement
Human tissue samples used in this study were obtained with written informed consent from the patients or their relatives. This study was approved by the Ethics Committee of Qilu Hospital of Shandong University.

Specimen collection
The study material includes 35 HCC and 35 adjacent normal tissues cases from Qilu Hospital of Shandong University, treated between 2015 and 2016. All patients underwent surgery only or postoperative adjuvant therapy. The patients' clinicopathologic information was obtained from the pathology department archive and was confirmed by hospital medical records. All patients consent forms were obtained prior to our study. Those who had received preoperative chemotherapy, radiotherapy, and embolization will be excluded. All HCC patients were classified based on the criteria of American Joint Committee on Cancer Staging Manual, Seventh Edition (2010).

Wound healing assay
HepG2 and Hep3B cells (3×10 5 cells/well) were seeded in six-well plates and incubated overnight. After cells had reached 80-90% confluence, the culture medium was removed, and a denuded area of constant width was created by mechanical scraping with a sterile 200-μl pipette tip. Cell debris was removed, and the monolayer was washed three times with serum-free medium. The cells were cultured for an additional 24 h, and wound closure was monitored and photographed under a microscope. Results are expressed as migration indices.

Cell migration assays
Migratory profiles were evaluated using Costar 24well plates containing polycarbonate filter inserts with 8-mm pores. HepG2 and Hep3Bs (3×10 5 cells/well) were added to the upper inserts and resuspended in serum-free medium. After 24 h, non-migrating cells were removed, and the underside of the insert was stained with crystal violet. Cells that had migrated into the lower chamber medium were collected and counted. Migration was calculated in terms of total cells on the lower membrane surface.

RNA interference
The Stat5 and control siRNAs (human STAT5-siRNA, sc-29495; control-siRNA, sc-37007) were obtained from Santa Cruz Biotechnology. siRNAs were transfected into HepG2 and Hep3B cells with the siRNA Reagent System (sc-29528, Santa Cruz Biotechnology) at a final concentration of 80 nM for 48 h according to the manufacturer's protocol.

Quantitative real-time PCR
HepG2 and Hep3B cells (3×10 5 cells/well) were seeded in 12-well plates and incubated overnight. For inhibitory assays, the cells were stimulated with human IGF-1 (100 ng/ml) in the presence of the Stat5 inhibitor (30 μM) or medium for an additional 6 h. For RNA interference experiments, HepG2 and Hep3B cells (1×10 5 cells/well) were transfected with siRNA for 48 h and then treated with human IGF-1 (100 ng/ml), respectively, for an additional 6 h. Then, total RNA was extracted. Reverse transcription and real-time quantitative PCR (qRT-PCR) were performed as previously described, and the mRNA levels of target genes were normalized to those of GAPDH. The primer sequences used for PCR are presented in Table 1.

Western blotting
HepG2 and Hep3B cells (1×10 6 cells/well) were seeded in six-well plates and incubated overnight. Protein concentrations of the lysates were quantified by using the Bradford assay to ensure equal amounts of protein loaded per lane in SDS-PAGE. Cell lysates were separated by SDS-PAGE and transferred onto PVDF membranes (Millipore, Billerica, MA, USA). The membranes were blocked in 5% BSA containing 0.05% Tween-20 and probed with the appropriate primary antibodies overnight at 4°C. The membranes were washed three times for 10 min each and incubated with horseradish peroxidase-conjugated secondary antibodies for 1 h at room temperature. Membranes were then washed as above and visualized by enhanced chemiluminescence.
All the antibodies were purchased from Cell Signaling Technology (CST, Beverly, MA).

Statistic analysis
The data are presented as the mean ± SEM. T-tests for independent events were used to evaluate the significance of differences. Data in different groups were compared by Kruskal-Wallis one-way analysis of variance. Correlations were used to analyze the relationships in HCC tissues. Statistical analyses were conducted using SPSS statistical software, version 19.0 (Chicago, IL); twotailed results with p<0.05 were regarded as significant.