14-3-3β Promotes Migration and Invasion of Human Hepatocellular Carcinoma Cells by Modulating Expression of MMP2 and MMP9 through PI3K/Akt/NF-κB Pathway

14-3-3β has been demonstrated to possess the oncogenic potential, and its increased expression has been detected in multiple types of carcinomas. However, majority of previous studies focused on the role of 14-3-3β in tumor cell proliferation and apoptosis, leaving much to be elucidated about its function in tumor cell invasion and metastasis. Hence, the present study aimed to investigate the role of 14-3-3β in the invasion of hepatocellular carcinoma (HCC) cells and the implications in the prognosis of HCC patients. We first examined the expression of 14-3-3β in the primary tumors of HCC patients with or without portal vein tumor thrombus (PVTT), and found that 14-3-3β expression was higher in the primary tumors with PVTT, and the level was even higher in the PVTTs. Kaplan-Meier curves and multivariate analysis revealed that high expression of 14-3-3β was associated with overall survival (OS) and time to recurrence (TTR) of HCC patients. In addition, ectopic expression of 14-3-3β in HCC cell lines led to enhanced migration ability and invasiveness, as well as up-regulation of matrix metalloproteinase 2 and 9, which could be suppressed by inhibiting the activation of Akt and nuclear factor-κB (NF-κB) signaling. Furthermore, we identified a correlated elevation of 14-3-3β and p-Akt in the primary tumors of HCC patients, and showed that a combinatory detection of 14-3-3β and p-Akt could better predict post-surgical outcome of HCC patients.


Introduction
Hepatocellular carcinoma (HCC) is one of most common cancers worldwide [1]. Despite the advanced modalities that are commonly applied in HCC patients, such as hepatic resection, liver transplantation, transcatheter arterial chemoembolization (TACE) and ablation therapy, the prognosis remains extremely poor [2]. The 5-year survival rate is less than 30% in HCC patients after surgical resection, mainly because of the high recurrence and metastasis rates.
However, the mechanisms underlying the recurrence and metastasis in HCC still remain unclear. Hence, further understanding of the underlying mechanisms is crucial for the development of novel therapeutic strategies, and would thereby improve the prognosis of HCC patients.
Accumulating evidences suggest that 14-3-3β plays an important role in tumorigenesis and tumor progression. For example, increased expression of 14-3-3β has been observed in a large number of solider tumors, including lung cancer [10], astrocytoma [11], glioma [12], squamous cell carcinoma [13], colorectal cancer [14], gastric cancer [15] and HCC [16]. Overexpression of 14-3-3β in NIH 3T3 cells has been identified to stimulate anchorage-independent growth and tumor formation in nude mice [17]. Reduction of 14-3-3β expression in rat hepatoma AFB1-K2 cells by forced expression of antisense 14-3-3β RNA significantly suppressed tumor cell proliferation and tumorigenesis [18], suggesting a pivotal role of 14-3-3β in the abnormal growth of tumor cells. Recently, Liu et al. reported that elevated 14-3-3β in HCC was a risk factor for extrahepatic metastasis and predicted a worse 5-year survival rate [16]. Nevertheless, the molecular mechanism underlying the role of 14-3-3β in HCC metastasis and progression remains elusive.
In the currently study, we found that high 14-3-3β protein expression in primary HCC tissues was associated with significantly worse clinical outcomes. The in vitro Transwell assays revealed that 14-3-3β promoted HCC cell migration and invasion. Mechanistically, 14-3-3β augmented the expression of matrix metalloproteinase 2 (MMP2) and MMP9 through PI3K/ Akt/NF-κB pathway, thereby enhancing the invasiveness of HCC cells. Furthermore, we show that a combinatory detection of 14-3-3β and p-Akt provides a better prognostic value for HCC patients. We have thus identified a novel pathway, PI3K/Akt/NF-κB/MMPs, which is activated by 14-3-3β in HCC malignancy.

Patients and samples
This study was reviewed and approved by the Clinical Research Ethics Committee of the General Hospital of Shenyang Military Area Command. Ninety-seven HCC patients who underwent curative resection in the General Hospital of Shenyang Military Area Command (Shenyang, China) from January2009 to March 2011 were randomly and retrospectively enrolled in this study in January 2013, and written informed consent was obtained from all participants. The researchers did not have access to the identifying information of the patients during or after date collection. HCC samples were retrieved to construct tissue microarray (TMA) to detect the expression of indicated proteins. All of the patients (100%, 97/97) had hepatitis B virus background. All patients were followed until December2014 with the longest follow-up up to 69 months. The time of the surgery was used to calculate the time to a given event. Overall survival (OS) and time to recurrence (TTR) were defined as previously described [19]. The diagnosis of tumor recurrence was based on cytologic/histologic evidence as well as the noninvasive diagnostic criteria for HCC used by the European Association for the Study of the Liver. Tumor stage was determined according to the 2002 International Union Against Cancer TNM classification system. Please see detailed clinicopathologic features in S2 Table. Microsatellite nodules that were defined as tumors adjacent to the border of the main tumor were only observed under the microscope. Microvascular invasion, defined as tumors spread to liver microvascualr but not to the main portal vein, was observed under the microscope. Early recurrence is defined as time to recurrence in less than 24 months. An additional 76 pairs of human HCC and peritumoral samples (collected between October 2011 and July 2012) were subjected to quantitative RT-PCR (64 pairs) and western blot analysis (12 pairs). Further, 34 pairs of HCC samples with portal vein tumor thrombi (PVTT) were obtained and used to perform quantitative RT-PCR (24 pairs) and western blot analysis (10 pairs).

Cell culture and transfection
Human hepatoma cell lines Hep3B, HepG2 and SMMC-7721 were purchased from the Cell Bank of Chinese Academy of Sciences (Shanghai, China). Hep3B and HepG2 cells were cultured in MEM (Gibco, Carlsbad, CA, USA) supplemented with 10% FBS (Hyclone, Logan, UT, USA), while SMMC-7721 cells were cultured in RPMI-1640 (Gibco) supplemented with 10% FBS. CSQT-2 cell line, which was derived from PVTT of HCC and established by Dr Cheng's lab [20], was a kind gift from Dr Cheng. CSQT-2 cells cultured in DMEM (Gibco) supplemented with 10% FBS. The cells were maintained at 37°C in a humidified atmosphere of 95% air and 5% CO 2 .

Western blot
For total protein extraction, tissue samples were physically homogenized and lysed with RIPA lysis buffer (Beyotime, Haimen, China). The transfected cells were incubated with the indicated inhibitors for 5 h and then lysed with NP-40 lysis buffer (Beyotime) to extract total proteins or were subjected to nuclear protein extraction with a Nuclear and Cytoplasmic Protein Extraction Kit (Beyotime) following the manufacturer's instructions. A total of 20 μg proteins from each sample were separated by SDS-PAGE, and transferred onto PVDF membranes (Millipore, Bedford, MA, USA). The membrane was incubated with a specific primary antibody against the protein of interest overnight at 4°C, followed by incubation with a horseradish peroxidase (HRP)-conjugated IgG secondary antibody (Beyotime) for 1 h at room temperature. Anti-14-3-3β antibody was purchased from ABGENT (San Diego, CA, USA); anti-p-IκB antibody was purchased from Bioss (Beijing, China); primary antibodies against MMP2, MMP9, P38, p-P38, JNK, p-JNK, ERK1/2, P-ERK1/2, Akt, p-Akt, P65, E-cadherin, N-cadherin and Vimentin were purchased from Wanleibio (Shenyang, China). The detailed information about the antibodies is shown in S1 Table. The immune complexes were finally visualized using the ECL system (7SeaPharmTech, Shanghai, China). To verify equal protein loading and transfer, membranes were stripped with the stripping buffer (Beyotime) and re-probed with anti-β-actin antibody (Wanleibio) for total proteins or with anti-Lamin A antibody (Santa Cruz, Dallas, TX, USA) for nuclear proteins.

Immunohistochemistry/Tissue microarray (TMA)
Matched pairs of primary HCC tissues and adjacent liver samples were used for the construction of a TMA (in collaboration with the Shanghai Biochip Company, Shanghai, China). The TMA construction was performed as previously described [21]. Immunostaining was performed on TMA slides following the routine protocol. Following antigen retrieval, the samples were incubated with 3% H 2 O 2 , blocked with goat serum and probed with anti-14-3-3β antibody (1:50;ABGENT) overnight at 4°C, followed by serial incubations with HRP-conjugated secondary antibody and diaminobezidine (Sigma-Aldrich) for chromogenic reactions. The samples were also probed for p-Akt, using anti-p-Akt antibody (1:50, Wanleibio), followed by reactions with alkaline phosphatase-conjugated secondary antibody and the chromogen substrate BCIP/NBT (Beyotime). Cell nuclei were counter-stained with hematoxylin.
The TMA slides were scanned with an Aperio ScanScope GL and assessed by the Aperio ImageScope software (Aperio Technologies, Vista, CA). Scoring of the TMA samples was based on the percentage of positively stained cells and the staining intensity. The scores equal to or above the median of all the values were defined as high, while the scores below the median were defined as low.

Transwell assay
Cell migration ability was assessed by the Transwell assay with uncoated upper chambers (Corning, New York, USA) [22], whereas cell invasion was examined by the Transwell assay with the upper chamber pre-coated with phenol red-free Matrigel (BD Biosciences, San Jose, CA, USA). The cells were pre-treated with 5 μM mitomycin-C for 2 h to inhibit proliferation. 2×10 4 cells resuspended in 200 μl serum-free medium were plated in the Transwell upper chamber, and allowed to migrate toward the bottom wells containing 800 μl culture medium supplemented with 20% FBS for 24 h in a 37°C incubator. Thereafter, cells remaining on the top surface of the Transwell microporous membrane were removed, and the cells on the bottom surface of the membrane were fixed with paraformaldehyde and stained with 0.1% crystal violate. Under a 200× inverted microscope, five fields on each membrane were randomly selected, and the mean number of invading cells was calculated.

Statistical analysis
The data were analyzed using the SPSS version 2.0 (IBM, Chicago, IL, USA). Wilcoxon signedrank test was used to analyze the differences between primary tumors and matched nontumorous tissues or PVTTs. The Student's t-test was used to compare differences between two groups. Kaplan-Meier curves of cumulative survival and recurrence were plotted, and the differences were analyzed by the log-rank test. Factors that influence post-surgical survival and recurrence were analyzed by multivariate analysis using Cox proportional hazard regression models, and the results are reported as hazard ratios (HR) with their 95% confidence intervals (CI). The correlation coefficients between the levels of 14-3-3β and p-Akt in the primary tumors were calculated by the Spearman correlation test. All P values were two-tailed, and the differences with p<0.05 were considered statistically significant.

Expression of 14-3-3β is up-regulated in human HCC tissues and invasive HCC cell lines
We first examined 14-3-3β expressionin paired HCC tissues. 14-3-3β transcripts were significantly increased in the primary tumor tissue of HCC relative to the paired nontumorous tissue in 64 HCC patients, and the elevation of 14-3-3β expression was further confirmed by western blot assay (S1A & S1B Fig). We next performed Immunohistochemical (IHC) analysis of 14-3-3β expression using HCC TMA containing 97 paired HCC samples. As shown in S1C Fig, the staining density of 14-3-3β protein in the tumor tissues was stronger than that in the nontumorous tissues (average staining density of 14-3-3β protein: 0.359±0.068 vs 0.267±0.041, p<0.001). Portal vein tumour thrombus (PVTT), arising from the invasion of HCC cells into the portal vein, serves as a poor prognostic factor of metastasis. The mRNA and protein levels of 14-3-3β were significantly increased in the PVTT compared with the matched primary tumor and the nontumorous tissue (Fig 1A & 1B). Furthermore, we examined the expression of 14-3-3β protein in three HCC cell lines (SMMC-7721, Hep3B and CSQT-2) and one normal liver cell line (HL-7702) by western blot analysis. As shown in Fig 1C, consistent with the results from the primary tissues, a higher level of 14-3-3β protein was observed in the more aggressive HCC cell line (CSQT-2) than that in the normal liver cells (HL-7702) and the less malignant HCC cell lines (SMMC-7721 and Hep3B). All of these results suggested a potential role of 14-3-3β in HCC metastasis.
The combination of 14-3-3β and p-Akt provides a better prognostic value for HCC patients As shown in Fig 5A & 5B, tissues microarray analysis of 97 patients' specimens revealed a significant positive correlation between 14-3-3β and p-Akt expression levels (p<0.01), which further supports the notion that 14-3-3β activates Akt in HCC. Elevation of either 14-3-3β or p-Akt in HCC tissues predicted a poor prognosis of the patient (S5 Fig). The patients bearing HCC with above-medium levels of both 14-3-3β and p-Akt suffered exacerbated adverse clinical outcomes compared with those with below-medium levels of both 14-3-3β and p-Akt (OS, p<0.001; TTR, p<0.001; Fig 5C & 5D). Collectively, our results suggest that combinatory evaluation of 14-3-3β expression and p-Akt signal is a power predictor of prognosis in HCC patients. In addition, the results reinforce the model of 14-3-3β-mediated activation of PI3K/ Akt/NF-κB pathway, which accounts for the up-regulation of MMP2 and MMP9 in HCC cells and thus promotes metastasis.

Discussion
Emerging evidences have been shown to support that 14-3-3 plays a critical role in the pathogenesis and metastasis of cancers. For example, 14-3-3z was found to interact with ErbB2 to regulate several important metastasis mediators including E-cadherin, fork head box protein M1 (FOXM1) and β-catenin, such as to promote breast cancer metastasis [23]. In addition, 14- 3-3ε has been demonstrated to promote EMT by inducing Zeb-1 and Snail expression, thereby promoting cell migration and invasion of HCC [24]. Liu et al. also identified that aberrant expression of 14-3-3β is associated with extrahepatic metastasis and worsened survival in HCC patients [16]. In the current study, consistent with Liu et al.'s findings, our results indicate that high 14-3-3β expression in primary HCC tissues is associated with intraheaptic metastasis and significantly worsened clinical outcome. Furthermore, we demonstrate that the pro-metastatic activity of 14-3-3β is most likely attributed to 14-3-3β-mediated activation of PI3K/Akt/NF-κB pathway, which consequently up-regulates the expression of MMP2 and MMP9 in HCC cells.
An earlier study reported that Snail can selectively interacted with14-3-3γ, 14-3-3ε, 14-3-3τ, 14-3-3η and 14-3-3β in MCF breast cancer cells and 293T cells, and such interaction represses the expression of E-cadherin and further triggers EMT [25]. However, in our study, elevated 14-3-3β Promotes HCC Invasion through PI3K/Akt/NF-κB 14-3-3β expression did not affect the expression of the key EMT-related markers in HCC cells, implying that14-3-3β-mediated HCC cell invasion and metastasis is EMT independent. The discrepancy between our results and the previous study is probably due to the differential functions that 14-3-3 exerts in different cell types.
The expression level of MMPs is implicated to be correlated with the metastatic ability of cancer cells. Particularly, abnormal expression of MMP2 and MMP9 is often detected in solider tumor tissues and is associated with tumor metastasis in many cancers including HCC [26]. In this study, we demonstrated that 14-3-3β overexpression in HCC cell line enhanced cell migration and invasion, accompanied by up-regulated expression of MMP2 and MMP9, suggesting that 14-3-3β-regulated expression of MMP2 and MMP9 is important for HCC invasion and metastasis. Previous functional studies of 14-3-3β mainly focused on its role in cell proliferation and tumor growth [16][17][18], and our study supports Liu's work [16] by demonstrating the important role of 14-3-3β in tumor invasion and metastasis. Furthermore, by using specific inhibitors of various cancer-related signaling pathways, we showed that 14-3-3β-induced upregulation of MMP2 and MMP9 was dependent on PI3K/Akt/NF-κB signaling pathway. Liu et al. previously demonstrated that 14-3-3β-promoted hepatic tumor cell migration and invasion could be suppressed by inhibiting MEK/ERK (MAPK) signaling pathway [16], which, however in our study, did not show any effect on 14-3-3β-induced up-regulation of MMP2 and MMP9.
It is generally believed that 14-3-3 proteins bind to a wide range of signaling proteins, especially kinases, facilitate their phosphorylation and protect their phosphorylated sites from dephosphorylation, thereby prolonging their activation status. Thus, enhanced activation of various signaling pathways was observed in 14-3-3β-overexpressing cells in our study. Raf proteins, which are the key effectors of Ras GTPase that initiates the activation of MEK/ERK cascade, were among the first identified 14-3-3 targets [27]. In addition to MEK/ERK cascade, the involvement of which in 14-3-3β-regulated tumor cell invasion has been described previously [16], we identified that PI3K/Akt and NF-κB signaling pathways also played a role in 14-3-3βregulated invasiveness of HCC cells. Akt is a key kinase that generates 14-3-3 binding sites on a diverse array of target proteins [28], thus inhibition of PI3K/Akt signaling abolishes the interactions of 14-3-3 proteins with various targets and therefore abate their functions, in this case, transactivation of MMP2/9 and other components that are required for cell invasion. In this study, we found that PI3K/Akt signaling also contributed to the activation of NF-κB pathway, and the crosstalk between them might be mediated via the signaling transducers such as IκB kinase (IKK) [29].
Consistent with a previous study [16], our clinical data indicated that increased 14-3-3β expression was associated with poor survival and high recurrence in HCC patients after surgical resection, supporting the clinical value of 14-3-3β as a prognostic marker in HCC. Further, with the molecular basis of the participation of PI3K/Akt pathway in 14-3-3β-regulated tumor cell invasion as well as the tissue microarray data, we show that 14-3-3β in combination with p-Akt, the level of which is positively correlated with the expression of 14-3-3β in the primary HCC tumors, can better predict post-surgical outcome of HCC patients.
In conclusion, our study provides clinical and experimental evidence to support that 14-3-3β plays an important role in HCC invasion and metastasis. Our results also suggest that 14-3-3β promotes HCC cell invasion by up-regulating MMP2 and MMP9 via PI3K/Akt/NF-κB signaling pathway. In addition, we propose a combinatory detection of 14-3-3β and p-Akt for more accurate assessment of prognosis of HCC patients.