Background
Hepatocellular carcinoma (HCC), which accounts for 85–90% of primary liver cancers, is a common malignancy worldwide and the second leading cause of cancer-related mortality [
1]. Especially in China, where it is accompanied by a high infection rate of hepatitis B virus, the importance of this disease should not be underestimated [
2]. Advances in modern medicine have resulted in the development of techniques for the diagnosis and therapy of HCC [
3‐
5]. Surgical resection and liver transplantation remain the treatment of choice for HCC patients in the early stage; however, most patients are at an advanced stage at presentation. Despite the fact that research into the treatment of HCC has been ongoing for decades, the prognosis and survival of HCC patients remain disappointing because of recurrence and metastasis [
6,
7]. Moreover, the mechanism underlying HCC development remains unclear, although many molecular biomarkers involved in HCC have been identified. Therefore, elucidating the potential mechanisms underlying HCC occurrence and development is critical to identify effective treatments for this disease.
GA binding protein (GABP) transcription factor alpha subunit (GABPA) is a subunit of the obligate heteromeric E twenty-six (ETS) transcription factor GABP. It harbors a highly conserved ETS motif that acts as a DNA-binding motif [
8,
9], as well as a protein-protein interaction domain for binding to the GABP beta subunit [
10]. GABPA regulates a broad range of genes involved in embryonic development, innate and acquired immunity, myeloid and hematopoietic stem cell differentiation, cell cycle progression, and migratory properties, and plays a role in certain human diseases.
GABPA regulates the expression of genes involved in mitochondrial function, and its inactivation results in early embryonic lethality [
11]. In addition, GABPA conditional deletion in mouse embryonic fibroblasts markedly decreases Tfb1m expression and reduces mitochondrial mass and protein synthesis, ATP production, and oxygen consumption [
12]. Deficiency of GABPA leads to a profound defect in B cell development and a compromised humoral immune response, in addition to thymic developmental defects [
13].
GABPA is involved in the maintenance and differentiation of hematopoietic stem and progenitor cells by activating the transcription of DNA methyltransferases and histone acetylases [
14]. In addition, GABPA is required for myeloid differentiation through the activation of the integrin alpha M promoter [
15]. Yang et al. reported that GABPA is required for myeloid differentiation in part by regulating the transcriptional repressor Gfi-1 [
16].
GABPA plays a major direct role in cell cycle progression. Conditional deletion of GABPA in mouse embryonic fibroblasts (MEFs) causes G1/S cell cycle arrest [
17], and reduces the numbers of cells entering the cell cycle [
18]. GABPA regulates cell survival and cell cycle progression through Yes-associated protein [
19]. GABPA is activated in a cell cycle-dependent manner and regulates the expression of genes related to cell cycle progression [
20]. Perdomo-Sabogal et al. used chromatin immunoprecipitation (ChIP) and comparative genomic approaches to identify newly evolved GABPA binding sites in 17 genes associated with a series of human diseases [
21]. Furthermore, a previous study showed that GABPA plays an important role in human chronic myelogenous leukemia (CML) and affects imatinib sensitivity [
22]. GABPA is required for the entry of hematopoietic stem cells into the cell cycle through the regulation of PRKD2 [
23].
A previous study showed that ablation of GABPA weakens the migratory properties of vascular smooth muscle cells by modulating the expression of kinase interacting with stathmin (KIS), which affects the phosphorylation and activity of p27 [
18]. Odrowaz and Sharrocks confirmed that GABPA plays a complex role in controlling breast epithelial cell migration by directly affecting the expression of RAC2 and KIF20A [
24]. However, studies on the role of GABPA in human cancer are rare, and whether GABPA is involved in HCC cell invasion and migration remains unclear.
The loss of E-cadherin, a calcium-dependent cell-cell adhesion protein, is associated with tumor migration, invasion, and poor prognosis. Epithelial cells can acquire a fibroblastoid morphotype accompanied by the acquisition of invasive and metastatic abilities in response to E-cadherin downregulation. Several transcription factors including Snail, Slug, and Twist among others are involved in the repression of E-cadherin gene transcription and the induction of epithelial-mesenchymal transition (EMT). However, to the best of our knowledge, there are no studies addressing the relationship between GABPA and E-cadherin expression.
In the present study, stably overexpressing and silencing GABPA cell lines were established to examine the potential role of GABPA in the regulation of HCC cell migration and invasion. GABPA expression was detected in human paired HCC tissue samples by western blotting and real-time PCR, and GABPA function was tested in vitro and in vivo. Finally, we investigated the potential molecular mechanisms underlying the effect of GABPA on HCC cell migration.
Methods
Cell culture
Six common HCC cell lines, MHCC-97H, PLC, BEL-7402, SMMC-7721, Huh7, SK-Hep1, and LO2, a normal liver cell lines, were purchased from the cell bank of Shanghai Institute of Cell Biology (Shanghai, China). All cells were cultured in RPMI-1640 or DMEM (Invitrogen) mediums. All the mediums were added with 10% fetal bovine serum (FBS) (Hyclon) and 100 units/ml of penicillin and streptomycin (Sigma). Cell lines were cultured according to the manufacturer’s protocol. All the cell lines were grown at 37 °C, in a 5% CO2 atmosphere, and passaged every 2–4 days.
Clinical samples
All of the clinical samples were obtained from chronic liver disease biological sample bank, department of Hepatobiliary Surgery, Zhongshan Hospital Xiamen University. None of the patients has received neoadjuvant therapy before surgical resection. The ethical approval was granted from the Committees for Ethical Review at the hospital. Written informed consent was also obtained from all patients based on the Declaration of Helsinki. The post-surgical patients were followed-up until September 2016.
Lentivirus vector based shRNA and overexpression
The pSIREN-RetroQ-puro RNA interference vector, which contained an RNA interference sequence that targeted GABPA or E-cadherin, was constructed similarly to the previous description [
25]. Forward and reverse short-hairpin RNAs (ShRNAs) which targeted GABPA or E-cadherin were annealed together respectively and inserted into the downstream from the promoter, finally generating the shRNA plasmid. The shRNA sequences were shown in Table
1. For GABPA over-expression plasmid, 1365 bp genomic sequence of GABPA coding region was cloned into the backbone of PBOBI-CMV vector downstream from the CMV promoter. The above mentioned plasmids and the virus packaging plasmids pMD2.G and PAX2 were transfected using the turbofect Transfection Reagent (Thermo, Cat #R0531) according to the manufacturer’s instructions. Then the HCC cells were transfected with virus-containing supernatant fluid and polybrene (10 μg/ml). Puromycin (2 μg/ml) was used for selection. Stable transfectants were maintained in conditional mediums with puromycin (1.0 μg/ml) for further analysis.
Table 1
Primers sequences
RT-PCR | GABPA | AAGAACGCCTTGGGATACCCT | GTGAGGTCTATATCGGTCATGCT |
E-cadherin | CGACCCAACCCAAGAATCTATC | AGGTGGTCACTTGGTCTTTATTC |
β-actin | ATAGCACAGCCTGGATAGCAACGTAC | CACCTTCTACAATGAGCTGCGTGTG |
shRNA | GABPA-1 | CCGGTGTTATCAGTAAGAAGTTCTAGCTTCAAGAGAGCTAGAACTTCTTACTGATAATTTTTTG | AATTCAAAAAATTATCAGTAAGAAGTTCTAGCTCTCTTGAAGCTAGAACTTCTTACTGATAACA |
GABPA-2 | CCGGTGATCTGGATCAATAACAACCTCTTCAAGAGAGAGGTTGTTATTGATCCAGATTTTTTTG | AATTCAAAAAAATCTGGATCAATAACAACCTCTCTCTTGAAGAGGTTGTTATTGATCCAGATCA |
E-cadherin | GATCCGCACCAAAGTCACGCTGAATTTCAAGAGAATTCAGCGTGACTTTGGTGTTTTTTACGCGTG | AATTCACGCGTAAAAAACACCAAAGTCACGCTGAATTCTCTTGAAATTCAGCGTGACTTTGGTGCG |
Pbobi-cmv | GABPA | GACTCTAGAGGATCCATGTACCCATACGACGTCCCAGACTACGCTACTAAAAGAGAAGCAGAGGAGC | AATTAATTCCTCGAGTTAATTATCCTTTTCCGTTTGCAGAGAAGC |
ChIP RT-PCR | E-Cadherin-P1 | CAGTTGCTATGATGAGCCAAGA | GGGAAGTCAGTGTTCTCCTTTG |
E-Cadherin-P2 | CTCTCATTGGCCTCAATCTCTC | GCCACTGACCAGCTCATTTA |
E-Cadherin-P3 | ACCACGCCTGGCTAATTT | GATCACGAGGTCAGGAGATTG |
E-Cadherin-P4 | CTCACTAACCCATGAAGCTCTAC | GCCGAGGCTGATCTCAAAT |
E-Cadherin-P5 | CACCTGTACTCCCAGCTACTA | GGTCTCACTCTTTCACCCAAG |
Western blot
Cultured cells were washed twice with ice-cold phosphatebuffered saline (PBS), then solubilized in a lysis buffer containing 1 mmol/L protease inhibitor cocktail (Sigma, St Louis, MO, USA) and quantified using the Bradford method. Protein lysate was separated by 6–12% sodium dodecyl sulfate polyacrylamide gel electrophoresis and then transferred to polyvinylidene difluoride membranes (Millipore, Billerica, MA, USA). After blocking the membranes with 0.05 g/mL non-fat milk, the blots were incubated with primary antibodies directly against GABPA (1:500, 21,542–1-AP, Proteintech), β-Actin (1:1000, #3700, CST), EZH2 (1:1000, #5246, CST) and E-cadherin (1:1000, 14472S, CST) at 4 °C overnight. Thereafter, the membranes were washed and incubated for 2–3 h at room temperature with the horseradish peroxidase conjugated secondary antibody. Protein bands were visualized with an enhanced chemiluminescence Reagent (K12045-D50, Advansta, USA) and quantified by densitometry via the Image-J software. The relative protein levels were calculated by comparing to the amount of β-Actin protein. Experiments were repeated in triplicate.
RT-PCR
Total RNA was extracted from tissues samples or cells using the Trizol reagent (Ambion, Cat 15,596–026, USA) according to the manufacturer’s instructions and then quantified at 260 nm using a NanoDrop 2000 spectrophotometer (Thermo Scientific, Waltham, MA, USA). Primers were designed and synthesized by BGI-Tech (Shenzhen, China). The sequences of the primer pairs were showed in Table
1. Total RNA (2 μg) was reverse-transcribed to complementary DNA (cDNA) using an RT kit (Promega, Madison, WI, USA). And quantitative PCR was performed in triplicate using Platinum SYBR Green qPCR Super Mix-UDG reagents (Invitrogen, Carlsbad, CA, USA) on a CFX96 Touch™ sequence detection system (Bio-Rad, Hercules, CA, USA). A dissociation procedure was performed to generate a melting curve for confirmation of amplification specificity. GAPDH was used as the endogenous control, and the comparative threshold cycle (2-ΔΔCT) equation was used to calculate the relative expression levels. All above were performed following the MIQE guidelines reported in the previous research [
26].
Wound healing assay
One day before the wound healing assay performed, HCC cells were seeded in 6-well plates. Once cellular density reached nearly 100% density, cells were scraped in a straight line with a 200 μl yellow micro-pipette tip, and photographed using phase-contrast microscopy to get the original width. Then the cells were put back into incubator. In order to assess migration distance, micrographs were taken every 12 h and quantified the difference between the original width and the width after cell migration. All assays were carried out in triplicates independently.
Chromatin immunoprecipitation
The chromatin immunoprecipitation (ChIP) assay was performed using an EZ-ChIP kit (Millipore, Catalog No. 17–10,461) according to manufacturer’s instructions. The E-cadherin promoter region located −3000 to −1 bp upstream of the transcription start site was amplified. Products were quantified by Real-time PCR method using both the ChIP-enriched DNA and input DNA as template. Enrichment by ChIP was assessed relative to the input DNA and normalized to the level of β-actin. The PCR primers for E-cadherin are listed in Table
1.
Migration and invasion assay
8-μm pore polycarbonate membrane inserts (Becton Dickinson, Franklin Lakes, NJ, USA) were used to measure the HCC cells’ invasive and migration abilities according to the manufacturer’ s protocol. In short, 2 × 106 cells in 250 μL serum free medium were seeded into the upper chamber and 500 μL medium containing 10% FBS was added to the lower chamber. After 48 h in culture, cells on the upper side were removed by a swab, fixed in 100% methanol for 15 min at room temperature, and then stained by crystal violet. Photographs of five random fields under 200 × magnification were captured for quantification analysis with the double-blind method. Three identical replicates were performed and eventually got a mean values.
Hematoxylin-eosin stained and immunohistochemistry
Tissues were fixed in 10% neutral formalin and then embedded in paraffin. 4 μm thick sections were prepared by pathological technologist. Hematoxylin-eosin (HE) stain was performed as previous described [
27]. For immunohistochemistry (IHC) staining, sections were deparaffinized, rehydrated, and then prepared for antigen retrieval and soaked in 3% H
2O
2 for 15 min at room temperature. Subsequently, the above sections were blocked with goat non-specific serum and incubated with GABPA antibody (1:400, 21,542–1-AP, Proteintech) and E-cadherin (1:100, 14472S, CST) at 4 °C overnight and biotin-labeled secondary antibody for 20 min at room temperature. Lastly, the sections were developed by dropwise adding DAB and stained with hematoxylin (Maixin Inc., Fuzhou, China). Evaluation of GABPA and E-cadherin staining in HCC tissue sections was performed refer to the IHC assessment methods used by Motoyuki Hashiguchi et al. previously [
28].
Animal assay
Male nude mice (4 to 5 weeks old) used in our study were purchased from Xiamen University and housed in Xiamen University laboratory animal center under pathogen-free conditions according to the institutional guidelines for animal care. All animal experiments met the National Institutes of Health Guidelines and were approved by the Committee on the Ethics of Animal Experiments of Xiamen University. As previous described [
29], mice were randomly assigned into two groups (10 cases for 7402-shCtrl group and 7402-ShGABPA group, respectively). 1.5 × 10
6 cells were re-suspended in PBS medium and then injected into the subcutaneous of armpit. The mice were sacrificed 40 days later and their lung and liver tissues were collected for metastatic foci examination via pathological stain.
Statistical analysis
Statistical analyses were performed using SPSS 21.0 (IBM, Chicago, IL, USA) and GraphPad Prism 5.0 (La Jolla, CA, USA) software. The results were expressed as the mean ± SD. Quantitative data were performed by two-related samples Wilcoxon non-parametric test for comparing the difference between two different groups. Categorical data were analyzed by X2 Test. Kaplan Meier analysis was used to evaluate the survival difference between subgroups. And the Spearman’s rank correlation analysis was used to examine possible correlations between GABPA and E-cadherin expression. P value less than 0.05 was considered as statistical significant.
Discussion
HCC is a common malignancy and its incidence and mortality are increasing worldwide. Despite advances in the surgical and medical treatment of HCC and extensive research into the mechanisms underlying HCC metastasis, the mortality from HCC remains high and the prognosis of patients is poor [
31]. Tumor invasion, metastatic dissemination, and recurrence are the major causes of the poor clinical outcome of HCC patients [
32‐
34]. Therefore, it is vital to identify metastasis-associated biomarkers and elucidate the mechanisms underlying HCC metastasis to develop effective therapeutic strategies to improve the quality of life of HCC patients.
Studies indicate that GABPA is involved in embryonic development, innate and acquired immunity, myeloid and hematopoietic stem cell differentiation, and cell cycle progression among other functions. However, to the best of our knowledge, the regulatory roles and mechanisms of GABPA in HCC cell migration and its clinical pathological significance have not been reported to date.
In the present study, we used a series of techniques to validate the role of GABPA in HCC metastasis and examined the potential underlying mechanisms. We demonstrated for the first time that GABPA is negatively associated with HCC progression, as indicated by the following results: First, GABPA protein and mRNA were downregulated in human HCC tissues compared with adjacent noncancerous tissues. GABPA was also consistently downregulated in HCC cell lines. Second, in vitro experiments showed that shGABPA vector-mediated GABPA knockdown in HCC cell lines markedly promoted cell migration and invasion, whereas ectopic expression of GABPA had the opposite effects. Third, in vivo assays in a xenograft model confirmed these results, as the number of metastatic tumors in the lungs and liver was higher in the sh-GABPA group than in the control group. Consistent with the tumor suppressive role of GABPA in cells, GABPA downregulation was associated with aggressive clinicopathological characteristics in HCC patients. Moreover, a low GABPA mRNA level was significantly associated with decreased survival time and worse prognosis in HCC patients.
A previous study indicated that GABPA exerts paradoxical roles in regulating cell migration and invasion. GABPA inhibits the migratory properties of vascular smooth muscle cells by controlling the expression of the kinase KIS [
18]. Odrowaz and Sharrocks reported that GABPA plays a complex role in controlling breast epithelial cell migration by directly affecting the expression of the RAC2 and KIF20A genes [
24]. Therefore, the effect of GABPA on cancer cell migration and invasion may be cancer and tissue specific and may involve different signaling pathways in different cells.
In the present study, we investigated the potential mechanisms underlying the effect of GABPA on HCC cell migration and invasion. Previous studies showed that EMT plays a pivotal role in HCC tumor metastasis [
35,
36] and contributes to early stage dissemination of cancer cells [
37]. We therefore examined the effect of GABPA on EMT markers and showed that GABPA knockdown downregulated E-cadherin at the protein and mRNA levels. However, overexpression of GABPA had a moderate effect on the mRNA and protein expression of E-cadherin. We speculated that shRNA-mediated GABPA knockdown led to a lower amount of GABPB indirectly binding to the E-cadherin promoter and inactivating its transcription, whereas overexpression of GABPA could not enhance the transcription of the E-cadherin gene significantly because of the limitation of the amount of GABPB, which forms a functional GABP transcription factor complex with GABPA. However, future studies are needed to verify these hypotheses.
GABPA was previously shown to physically interact with methyltransferase like 23 to regulate thrombopoietin and ATP5B gene expression [
38]. Additionally, Lucas et al. showed that GABPA was selectively enriched at HS2 in human cells, and its occupancy was inversely correlated with CpG island methylation of the TMS1 gene [
39]. These results suggest the epigenetic regulation of GABPA expression. In the present study, we accidentally found that EZH2 was negatively regulated by GABPA (Additional file
4: Figure S4a). Han et al. reported that EZH2 suppresses E-cadherin expression and promotes pancreatic cancer cell migration [
40]. Therefore, we speculated that GABPA may indirectly promote E-cadherin expression because of its effect on EZH2. To test this hypothesis, the expression level of E-cadherin was detected by western blotting in three groups of cells: Huh7-Control, Huh7-GABPA, and Huh7 cells overexpressing GABPA and EZH2. As shown in Fig. S4b, upregulation of GABPA increased E-cadherin protein levels, and this effect was moderately restored by EZH2. These results demonstrate that GABPA regulates E-cadherin via EZH2. However, the specific mechanism needs further clarification.
Acknowledgements
We sincerely appreciate all of the patients participated in this study. We also thank Cheng-rong Xie and Jie Li for the valuable assistance with ChIP assay and data analysis.