STYK1 promotes epithelial-mesenchymal transition and tumor metastasis in human hepatocellular carcinoma through MEK/ERK and PI3K/AKT signaling

Serine/threonine/tyrosine kinase 1 (STYK1) is known to be involved in tumor progression. However, its molecular role and mechanism in hepatocellular carcinoma (HCC) remains unknown. We evaluated the effect of STYK1 expression in HCC tissues and investigated the underlying mechanisms associated with progression. HCC tissues expressed greater levels of STYK1 than paired non-tumor tissues. Patients with HCC expressing low levels of STYK1 showed both, greater disease-free (p < 0.0001) and overall (p = 0.0004) survival than those expressing high levels of STYK1. Decreased expression of STYK1 was significantly associated with decreased cell proliferation, reduced migratory capability, and reduced invasive capability. Overexpression of STYK1 was significantly associated with increased cell proliferation, migratory capability, and invasive capability in vitro, as well as increased volume of tumor, weight of tumor, and number of pulmonary metastases in vivo. Furthermore, STYK1’s mechanism of promoting cancer cell mobility and epithelial-mesenchymal transition (EMT) was found to be via the MEK/ERK and PI3K/AKT pathways, resulting in increased expression of mesenchymal protein markers: snail, fibronectin, and vimentin, and decreased E-cadherin expression. Our results suggest that STYK1 acts as an oncogene by inducing cell invasion and EMT via the MEK/ERK and PI3K/AKT signaling pathways and it therefore may be a potential therapeutic target in HCC.

Accumulating evidence has shown that epithelial-to-mesenchymal transition (EMT), characterized by the progression of epithelial phenotype to mesenchymal phenotype, plays a critical role in cancer invasion and metastasis during tumor progression [16][17][18] . EMT-associated genes are reported to be associated with HCC recurrence and metastasis [19][20][21] . In HCC, EMT results in increased venous invasion, metastasis, and a poorer prognosis 5,20,21 . Despite evidence of STYK1 acting as an oncogene and being involved in a variety of cancers, the role of STYK1 in regulating HCC metastasis and EMT remains unclear.
In this study, we estimated STYK1 expression and its effects on invasion and metastasis of HCC. We also investigated the precise mechanism/pathway by which STYK1 contributes to invasion and EMT events in vitro and in vivo. These findings may provide a potential therapeutic target to inhibit HCC progression and metastasis.

Results
STYK1 levels are elevated in HCC tissue compared to paired non-tumor tissue. Immunohistochemical staining revealed the presence of STYK1 at significant levels in HCC tissues and relatively insignificant levels in paired, non-tumor tissues (Fig. 1A). Then we detected the expression of STYK1 in human normal liver cell line LO2 and 5 HCC cell lines (7402, 7721, Hep3B, 97H, and LM3), and results demonstrated a higher expression of STYK1 in HCC cell lines, compared with LO2 (Fig. 1B). These results suggested that STYK1 might play an important role in HCC progression and development.

STYK1 level functions as an independent predictor of survival in patients with HCC.
To further explore the potential role of STYK1 in HCC progress, we evaluated STYK1 expression and the clinicopathological characteristics are summarized in Table 1. The patients were divided to two groups (low STYK1 and high STYK1) according to median of STYK1 expression in cancer tissues samples. STYK1 overexpression was significantly correlated with tumor size, vascular invasion, and a higher tumor-nodule-metastasis (TNM) stage. Patients with HCC tumors expressing low STYK1 levels had significantly higher overall survival (OS) than those expressing high STYK1 levels (p = 0.0004) (Fig. 1C). Similarly, patients with HCC tumors expressing low STYK1 levels showed significantly longer disease-free survival (DFS) than those expressing high STYK1 levels (p < 0.0001) (Fig. 1D). Moreover, tumor size, tumor number, vascular invasion, and advanced TNM stage were found associated with lower OS and shortened time-to-recurrence (TTR) in univariate analysis (Table 2). Multivariate Cox regression analysis further revealed that STYK1 was an independent risk factor for lower OS and shortened TTR in patients with HCC (Table 3).

STYK1 silencing inhibits HCC cell growth, migration, and invasion.
To explore the role of STYK1 in the progression of HCC, short hairpin RNA (shRNA) were used to silence endogenous expression of 2 HCC cell lines: LM3 and 97H, which show high malignant and metastatic capacity. Western blot was used to confirm decreased expression of STYK1 in LM3 and 97H cell lines following transfection with negative vector (SCR) and different knockdown vectors. Sh3, targeting STYK1, exhibited significantly decreased expression of STYK1 as compared to SCR, sh1, and sh2; therefore it was selected for subsequent studies ( Supplementary Fig. 1). MTT assay revealed that the proliferation rate of LM3 and 97H cells transfected with STYK1-knockdown was lower than that of negative vector cells ( Fig. 2A). A colony-formation assay was further performed to assess cell proliferation. Similarly, LM3 and 97H cells with STYK1 silencing exhibited significantly smaller and lower number of colonies (Fig. 2B). In addition, migratory and invasive capacities were assessed using a transwell assay. As shown in Fig. 2C, inhibition of STYK1 expression resulted in decreased cell migration and invasion in both, LM3 and 97H cells. In summary, our results indicate that down-regulation of STYK1 decreased the proliferation and metastatic ability of HCC in vitro.  Fig. 2). Ectopic STYK1 expression significantly promoted proliferation activity in both, 7721 and 7402 cells (Fig. 3A). In addition, compared with control vector, the colony-formation assay showed that the size and number of colonies formed by 7721 and 7402 cells were significantly increased by stable STYK1 expression (Fig. 3B). Furthermore, to assess the effects of STYK1 on HCC growth in vivo, 7402 cells with control vector or STYK1 were injected subcutaneously into nude mice. The tumors size was significantly larger in the STYK1-expression group compared with the control group (Fig. 3C). The growth curve and weight of dissected tumor further confirmed the progress of tumor cell growth caused by STYK1 overexpression (Fig. 3D). IHC also indicated more Ki67 antigen-positive cells in tumors derived from 7402 cells with STYK1 compared with tumors derived from vector (Fig. 3E). These data indicate that STYK1 played the role of an oncogene in HCC progression.

Overexpression of STYK1 promotes cell migration and invasion in vitro and in vivo.
The invasive and migratory capacities of HCC cells with STYK1 vector were assessed using a transwell assay. As expected, ectopically expressed STYK1 in 7721 and 7402 cells greatly increased migration and invasion abilities, compared with those shown by control vector cells (Fig. 4A,B). We then investigated if STYK1 enhanced cancer cell metastasis in vivo. As shown in Fig. 4C, ectopic expression of STYK1 resulted in increased metastatic foci and larger nodules in the lung. The results indicate that overexpression of STYK1 in 7402 cells led to significant promotion of cell invasion and lung metastasis in vitro and in vivo.

STYK1 induces epithelial-mesenchymal transition in HCC cells.
Previous studies showed that cell migration and invasion are associated with altered levels of EMT biomarkers 17 . To determine the underlying mechanisms of STYK1 in EMT, western blot analysis was performed to detect the expression of epithelial and mesenchymal protein markers. Our results showed that STYK1 up-regulated the expression of mesenchymal    5A). Conversely, STYK1-depleted LM3 and 97H cells exhibited increased E-cadherin expression, but decreased vimentin, fibronectin, and snail expression (Fig. 5B). MEK/ERK and PI3K/AKT signaling pathway are known to play critical roles in tumor invasion and metastasis. We estimated the alteration of ERK and AKT. As shown in Fig. 5A, overexpression of STYK1 promoted phosphorylation of ERK and AKT, while silencing STYK1 reversed this phenotype (Fig. 5B). To determine whether STYK1 promoted the invasiveness of HCC via EMT, immunofluorescent staining was performed. As shown in Fig. 5C, STYK1 induced vimentin staining in both, 7721 and 7402 HCC cells. To confirm these findings in vivo, immunohistochemical analysis was used to determine the expression of epithelial and mesenchymal protein. Consistently, we found that STYK1 overexpression displayed a decrease in the E-cadherin level but an increase in vimentin level (Fig. 5D), which represents the EMT phenotype in the mice bearing the 7402 xenografts. To further confirm the data, the expression of E-cadherin and vimentin were detected by IHC in the same HCC tissues samples. The result showed that E-cadherin was decreased, but vimentin was increased in HCC tissues (Fig. 5E), which consisted with the in vitro results. Overall, these results suggest that STYK1 promoted tumor invasion and metastasis by inducing EMT.  (Fig. 6A,B). In addition, inhibition of ERK1/2 by U0126 or of AKT by MK2206 reversed the effects on vimentin, fibronectin, twist, and snail up-regulation and E-cadherin down-regulation induced by STYK1 (Fig. 6C). To assess whether regulation of the PI3K/AKT and MEK/ERK signaling by STYK1 could be recapitulated in vivo, the tissue lysates of tumors from 4 mice were used to determine the expression of AKT, p-AKT, ERK, p-ERK, E-cadherin, vimentin, and twist by western blotting. Our data suggested overexpression of STYK1 increased the expression of p-AKT, p-ERK, vimentin and twist and decreased the expression of E-cadherin (Fig. 6D). In addition, treatment of nude mice with U0126 (inhibitors of MEK1/2) or MK2206 (inhibitors of AKT) significantly reduced the metastatic nodules induced by STYK1in lung (Fig. 6E,F). These results indicate that STYK1 may induce cell migration, invasion, and EMT through MEK/ERK and PI3K/AKT signaling pathways.

Discussion
A growing body of evidence has identified genes specifically upregulated or downregulated in HCC tissues that can be considered as early diagnostic markers, prognostic markers, and therapeutic targets [22][23][24][25] . STYK1 is one such a gene that is highly up-regulated in various types of tumors [9][10][11][12]14 ; however, little is known about its expression in HCC, with even less clarity on its underlying mechanisms. In our present study we found that STYK1 expression was significantly upregulated in HCC tissues. Further correlation analyses suggested that STYK1 overexpression correlated with tumor size, vascular invasion, and higher TNM stage. A multivariable Cox's regression analysis indicated that STYK1 was a significant risk factor of OS and DFS. In addition, depletion of endogenous STYK1 expression in LM3 and 97H cells suppressed cell growth, migration, and invasion, whereas overexpression of STYK1 in 7721 and 7402 cells had the opposite effect. Moreover, we found that STYK1 induced EMT via the MEK/ERK and PI3K/AKT pathway, which may be responsible for its invasion-promoting activity. Uncontrolled growth and metastasis, which are important features of malignant tumors, are the main cause of cancer-related death. STYK1 has been reported to promote BaF3 cell proliferation and invasion in vitro and in vivo 7 . Meanwhile, knockdown of STYK1 decreased proliferation of prostate cancer cell viability in vitro 14 .
Here, we demonstrated that STYK1 knockdown inhibited cell proliferation and invasion, while overexpression of STYK1 promoted this effect both in vitro and in vivo, suggesting the oncogenic role of STYK1 in HCC, consistent with its role in other cancers 7,13 . Accumulated data show that EMT results in increasing cell migration and invasion in several cancers 26,27 . We found that STYK1 overexpression induced EMT by elevating expression of the mesenchymal markers, vimentin, fibronectin, and snail and by reducing expression of the epithelial marker, E-cadherin. In contrast, increased expression of E-cadherin accompanied by decreased expression of vimentin, fibronectin, and snail were observed in LM3 and 97H with depleted STYK1.
Previous reports show that activation of MEK/ERK and PI3K/AKT contributes to cell growth 28,29 , promotes invasion, and EMT 30,31 . We showed that phosphorylated levels of ERK and AKT were enhanced by STYK1 overexpression and inhibited by STYK1 silencing, which is consistent with previous reports 7 . Furthermore, we investigated if MEK/ERK and PI3K/AKT signaling are responsible for STYK1-mediated acceleration of tumor cell migration, invasion, and EMT. Blocking MEK/ERK and PI3K/AKT pathway using special inhibitors significantly attenuated STYK1-enhanced migration and invasion of HCC cells. More importantly, U0126 or MEK2206 treatment inhibited the activation of vimentin, fibronectin, and snail as well as inactivation of E-cadherin by STYK1. Inhibition of MEK/ERK and PI3K/AKT pathway with inhibitors significantly reduced the metastatic nodules induced by STYK1 in the lung. Therefore, our data support that activation of MEK/ERK and PI3K/AKT signaling pathways were required for STYK1-stimulated cell migration, invasion, and EMT of HCC cells. STYK1 was previously shown to cause GSK-3β (Ser9) phosphorylation via activating AKT phosphorylation at the Thr308 residue in cervical cancer Hela cell 15 . Several studies have shown aberrant activation of AKT and GSK-3β involved in HCC tumorigenesis, EMT and Metastasis 32,33 . The role of STYK1/AKT/GSK3 axis in HCC progress might extend the key functional pathways to abnormal proliferation invasion, and EMT of HCC cells. Transforming growth factor β (TGFβ ), a potent inducer of the EMT, have been involved in medicating actin filament reorganization and invasiveness through activating many key genes and signaling pathways, the relation among TGFβ , STYK1 and EMT needed to be further investigated.
In conclusion, we demonstrated that STYK1 is highly expressed in tumor tissues and is significantly correlated with poor outcome of HCC. In addition, STYK1 promoted proliferation, invasion, and metastasis in vitro and in vivo. We also emphasized that MEK/ERK and PI3K/AKT are required for STYK1-mediated HCC cell invasion and EMT.   Lentivirus was packaged in HEK293T cells using X-tremeGENE (Roche, Basel, Switzerland) and the viral DNA was transduced into HCC cell lines. Cells were selected with medium containing 2 μ g/ml Puromycin (Sigma) after 72 h of infection.
Immunohistochemistry and evaluation of staining results. Immunohistochemistry was performed as previously described 34 . Briefly, 4-μ m thick tissue sections were used and stained with primary, rabbit polyclonal antibodies against STYK1 (Abcam, Cambridge, MA, USA) at 1:100 dilution and incubated at 4 °C. The stained sections were assessed by 2 independent observers (X.C.Z and W.Z.W.), who were blinded to the clinicopathological information of the patients. STYK1 staining was scored based on the staining intensity and percentage of cell staining, as previously described 34 . Colony-formation assay. For colony-formation assays, cells were plated on 6-well plates at a concentration of 800 cells/well and the medium was replaced every 3 days. Cells were fixed and stained with 1% crystal violet after incubating for 2 weeks.
Cell proliferation, migration and invasion assays. Cell proliferation assay was performed as previously described 34 . Briefly, cells (3 × 10 3 /well) in 100 μ l medium were seeded in 96-well plates (Corning, NY, USA). MTT was added to the wells at different time points and cells were incubated at 37 °C for 4 h. Absorbance was measured at 570 nm using a Coulter Counter (Beckman Coulter, Fullerton, CA, USA). Invasion chamber assay with Matrigel (invasion) or without (migration) was performed in triplicate, as previously described 34 . In brief, for migration assay, approximately 3 × 10 4 cells in serum-free medium were plated into the upper chamber of an 8-μ m pore size insert without Matrigel (migration) or with Matrigel (invasion) in the 24-well plate (Corning) and cultured for another 24 h. Cells were fixed in 4% paraformaldehyde and stained with 0.5% crystal violet. The non-migrated/non-invaded cells on the surface of the upper membrane were removed with a cotton swab. The membranes were photographed at 100× magnification under a microscope and the number of migratory/invasive cells was calculated in five randomly selected fields.
Immunofluorescence. For immunofluorescence microscopy, the cells were seeded on cover slips and incu- Western blot analysis. Total protein from cells and tissues were lysed in RIPA buffer (CST, Danvers, MA, USA) supplemented with complete protease inhibitor cocktail (Roche, Basel, Switzerland) before use. Equal amount of protein lysates were subjected to 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride membranes (Millipore, Billerica, MA, USA). The membrane was blocked in 5% nonfat milk in phosphate-buffered saline supplemented with Tween 20 at room temperature for 1 h and incubated with primary antibody and subsequently with appropriate horseradish peroxidase-conjugated, secondary antibody (1:5000; Pierce, Rockford, IL, USA). Protein bands were visualized by enhanced chemiluminescence detection system (Pierce, Rockford, IL, USA). The primary antibodies against p-AKT (Ser473), AKT, p-ERK (Thr 202/Tyr 204), ERK, E-cadherin, Snail, and GAPDH were obtained from Cell Signaling Technology (Danvers, MA, USA). Fibronectin and Vimentin were obtained from Abcam (Cambrdige, MA, USA). TWIST1 Antibody was obtained from Proteintech (Chicago, IL, USA).
In vivo tumorigenesis and metastasis assays. Nude nu/nu mice (4-6 weeks old) were purchased from the Shanghai Slaccas Animal Center (Shanghai, China). All mice were housed in specific pathogen-free conditions following the guidelines of the Institutional Animal Care and Use Committee of Shanghai Jiao Tong University School of Medicine. To assess the effect of STYK1 on tumorigenicity in vivo, 7402 cells (1 × 10 7 cells) with vector or STYK1 were subcutaneously injected into the left flank of the mice (8 mice/group). Xenograft growth was monitored every week using a caliper. Tumor volume was calculated by the following formula: tumor volume = length/2 × (width/2) 2 . Immunohistochemical staining for STYK1, Vimentin, and E-cadherin was performed on sections from the tumor.
To produce experimental metastases, 7402 cells (1 × 10 6 ) with vector or STYK1 were harvested in serum-free DMEM medium and injected into the tail vein of nude mice (10 mice/group). For U0126 and MK-2206 treatment, 7402 cells (1 × 10 6 ) stably overexpression STYK1 were randomly injected into the tail vein of nude mice and divided into control and treatment groups. The control group received vehicle (1% DMSO in Saline solution) alone. A dose of 5 mg of the U0126 per kg body weight per day was injected intraperitoneally into mice on days 14 through 21. A dose of 120 mg of the MK-2206 per kg body weight (P.O) per day was administered to mice on days 14 through 21. The mice were sacrificed after 6 weeks and the lungs were surgically excised and stained with H&E. Lung metastatic lesions were evaluated under a dissecting microscope (× 200 magnification).
Statistical analysis. All data are presented as mean value ± standard deviation (SD) from at least 3 independent experiments. Statistical analysis was performed using Student's t-test (two-sided) or analysis of variance. Chi-square (x 2 ) or Fisher's exact test was used to evaluate the associations between the expression of STYK1 and clinicopathological parameters. Cumulative overall and DFS were assessed using Kaplan-Meier's method and the log-rank test. Statistical analysis was performed using SPSS 16.0 software (SPSS, Inc., Chicago, IL, USA) and p < 0.05 was considered to be statistically significant.