Background
HCC is the fifth most common malignant tumor worldwide, with an incidence of approximately 626,000 cases each year [
1,
2]. In China and Southeast Asia, HCC is highly associated with viral hepatitis B and cirrhosis [
3]. The prognosis of patients with HCC has largely improved due to extensive advances in surgical techniques and diagnostic methods in recent years. However, the long-term prognosis is still unsatisfactory, largely due to the high recurrence and invasion rates even after resection (50-70% at five years) [
4,
5]. Unfortunately, little is known with respect to the molecular mechanisms underlying this aggressive behavior. Therefore, there is great demand to find a reliable prognostic biomarker, which may help clinicians predict the characteristics of the malignancy and decrease the rate of unfavorable outcomes in a high-risk population.
ATAD2 (ATPase family AAA domain-containing protein 2), a member of the AAA + ATPase family of proteins, was identified by microarray analysis [
6], contains both a bromodomain and an ATPase domain, and maps to chromosome 8q24 in a region that is frequently amplified in cancer [
7]. The structure of ATAD2 suggests that it has functions related to genome regulation, including cell proliferation, differentiation and apoptosis. Studies have revealed that ATAD2 is highly expressed in several types of tumors such as breast cancer, lung cancer, and large B-cell lymphoma [
8‐
13]. And recently Huang Q et al. has reported that a novel, highly up-regulated exon-exon junction was detected in ATAD2 gene by RNA-seq and the gene was highly expressed in HCC tissues [
14]. However, there have been no studies on the gene function and prognosis associated with ATAD2 in HCC.
In the present study, we detected the expression of ATAD2 in HCC and matched adjacent non-cancerous liver tissues. Different methods were used to determine the relationships between the expression of ATAD2, its clinical relevance, and the overall survival (OS) after resection. In addition, the effects of ATAD2 expression on cell invasion and metastasis were investigated in SMMC7721, QGY-7701, Bel-7402, PLC5, Huh7, HCCLM3, HepG2, and LO2 cell lines using small interfering RNAs. Also the miR-372 was identified as a direct and functional target for ATAD2 in hepatic carcinogenesis. Therefore, both ATAD2 and miR-372 appear to be good targets to study in our research.
Methods
Patient tissue samples
Primary tumor samples were obtained from 129 patients (106 males and 23 females) who were diagnosed with HCC and had undergone routine hepatic resection in the First Affiliated Hospital of China Medical University between January 2006 and December 2009. The follow-up period for survivors was 5 years. None of the patients had received preoperative radiotherapy or chemotherapy prior to surgical resection. The histological diagnosis and differentiation were evaluated independently by two pathologists according to the WHO classification system [
15]. The clinicopathological features are shown in Table
1. Fresh specimens were snap-frozen in liquid nitrogen and stored at -70°C immediately after resection for the extraction of RNA and protein. The project protocol was approved by the Institutional Ethics Committee of China Medical University prior to the initiation of the study. All patients provided written informed consent for the use of the tumor tissues for clinical research.
Table 1
Distribution of ATAD2 and Clinicopathological Characteristics in HCC patients
Total cases | 129 | 83 | 46 | |
Age (years) | | | | |
≥60 | 54 | 34(63.0%) | 20(37.0%) | 0.782 |
<60 | 75 | 49(65.3%) | 26(34.7%) |
Gender | | | | |
Male | 106 | 69(65.1%) | 37(34.9%) | 0.701 |
Female | 23 | 14(60.9%) | 9(39.1%) |
Tumor size | | | | |
≥ 5cm | 83 | 53(63.9%) | 30(36.1%) | 0.877 |
< 5cm | 46 | 30(65.2%) | 16(34.8%) |
Metastasis | | | | |
Yes | 72 | 54(75%) | 18(25%) | 0.005 |
No | 57 | 29(50.9%) | 28(49.1%) |
HBsAg status | | | | |
Positive | 73 | 43(58.9%) | 30(41.1%) | 0.141 |
Negative | 56 | 40(71.4%) | 16(28.6%) |
Tumor differentiation | | | | |
WD | 54 | 34(63.0%) | 20(37.0%) | 0.955 |
MD | 60 | 39(65.0%) | 21(35.0%) |
PD | 15 | 10(66.7%) | 5(33.3%) |
Cirrhosis | | | | |
Yes | 86 | 58(67.4%) | 28(32.6%) | 0.298 |
No | 43 | 25(58.1%) | 18(41.9%) |
Serum AFP | | | | |
<200 ng/dl | 41 | 29(70.7%) | 12(29.3%) | 0.301 |
≥200 ng/dl | 88 | 54(61.4%) | 34(38.6%) |
Tumor stage | | | | |
TNM I+II | 42 | 25(59.5%) | 17(40.5%) | 0.427 |
TNM III+IV | 87 | 58(66.7%) | 29(33.3%) |
Liver cancer cell lines and cell cultures
The liver cancer cell lines, SMMC7721, QGY-7701, Bel-7402, PLC5, Huh7, HCCLM3, HepG2, and the normal liver cell line, LO2, were obtained from the Shanghai Cell Bank (Shanghai, China). SMMC7721, QGY-7701, Bel-7402 and PLC5 cells were cultured in RPMI-1640 (Invitrogen, Carlsbad, CA). Huh7, HCCLM3, HepG2 and LO2 cells were grown in DMEM (Invitrogen). All media were supplemented with 10% fetal calf serum (Invitrogen) and 100 IU/ml penicillin (Sigma, St. Louis, MO).
RNA preparation and quantitative real-time PCR
Total RNA was extracted according to the manufacturer’s instructions. RT-qPCR was performed using a SYBR Premix Ex Taq (TaKaRa) on a Thermal Cycler Dice Real Time System (TaKaRa) with the following protocol: 30 s at 95°C followed by two-step PCR for 40 cycles of 95°C for 5 s and 64°C for 30 s. Each reaction was performed as we previously described [
3]. The primer sequences were as follows: ATAD2 forward, 5′-GGAATCCCAAACCACTGGACA-3′;
ATAD2 reverse, 5′-GGTAGCGTCGTCGTAAAGCACA-3′;
GAPDH forward, 5′-ATAGCACAGCCTGGATAGCAACGTAC-3′;
GAPDH reverse, 5′-CACCTTCTACAATGAGCTGCGTGTG-3′.
Western blots
Total protein was extracted from tumor tissues, non-tumor adjacent tissues and liver cancer cell lines using the Nuclear and Cytoplasmic Protein Extraction Kit (Beyotime, China). Fifty micrograms of total nuclear protein was separated by SDS-PAGE and then electrotransferred to PVDF membranes (Millipore, Billerica, MA, USA). Milk (5%) was used to block membranes for 2 hours at room temperature. After blocking, primary antibodies, including ATAD2, APC, CTNNA1 rabbit polyclonal antibody and GAPDH mouse polyclonal antibody (both diluted at 1:1000, Sigma, St. Louis, MO, USA), were incubated with the membranes overnight at 4°C. The secondary antibodies were then incubated for 2 hours at room temperature. Protein bands were identified using an ECL system (Millipore, Bedford, MA, USA)
Immunohistochemistry
ATAD2 expression was analyzed on paraffin-embedded specimens from 129 patients. Tissue sections of 4-μm thickness were deparaffinized in xylene and dehydrated before antigen retrieval for 5 min with an autoclave. Hydrogen peroxide (0.3%) was used to block endogenous peroxidase activity; nonspecific immunoglobulin binding sites were blocked by normal goat serum for 30 min at 37°C. Tissue sections were incubated with anti-ATAD2 (1:200, Sigma) overnight at 4°C. Biotinylated goat anti-rabbit immunoglobulin G was used as a secondary antibody (1:300, Sigma). After washing, the sections were incubated with streptavidin-biotin conjugated with HRP for 30 min, and the peroxidase reaction was developed with 3, 3′-diaminobenzidine tetrahydrochloride (DAB).
Semi-quantitative assessment and scoring
ATAD2 expression levels were scored semiquantitatively according to the percent of positively stained cells combined with the staining intensity. Samples were considered positive for ATAD2 if the nucleus or cytoplasm of the sample cells presented a positive staining. The percent positivity was defined as “0” if 0%, “1” if 1-10%, “2” if 11-50%, “3” if 51-80%, and “4” if >80%. The staining intensity was scored as “0” (no staining), “1” (weakly stained), “2” (moderately stained) and “3” (strongly stained). Both the percent positivity and the staining intensity were assessed by two doubly blinded investigators. The ATAD2 expression score was calculated from the value of percent positivity score × the staining intensity score. This value thus ranged from 0 to 12, and the tumors were classified into the following: negative (-), score 0; lower expression (1+), score 1–4; moderate expression (2+), score 5–8; and strong expression (3+), score 9–12. The immunohistochemical ATAD2 staining was grouped into two categories: low expression (0/1+) and high expression (2+/3+).
Depletion of ATAD2 by small interfering RNAs
On-TargetPlus SMARTpool siRNA for ATAD2 (No. LU-017603-00-0002) and On-TargetPlus siControl (D-001810-01-20) were purchased from Dharmacon. For transfections, 24 h after the cells were seeded in a 6-well plate, they were transfected with ATAD2 siRNA or control siRNA using DharmaFECT 1 according to the manufacturer’s protocol. The mRNA and protein levels were detected 48 h later.
Cell cycle analysis
Huh7 and HCCLM3 cells in 6-well plates were transfected with ATAD2 siRNA or negative control siRNA. After 48 h of transfection, cells seeded at a density of 5 × 105 per well were trypsinized, fixed with 70% ethanol at 4°C, and washed with PBS. A quantity of 100 μL RNase A was added, and the mixture was incubated in a 37°C water bath for 30 min. An additional 400 μL PI staining solution was added and incubated at 4°C in the dark for 30 min; a computer was then used to detect and record the red fluorescence upon excitation at a wavelength of 488 nm.
Cells were plated in 96-well plates in media containing 10% FBS at approximately 2,000 cells per well, 24 h after transfection. Then, 10 μl of CCK8 (Thiazolyl Blue) solution was added to each well and incubated for 1 h at 37°C. The results were quantified spectrophotometrically using a test wavelength of 450 nm.
After transfection, logarithmic growth phase cells in monolayer culture were prepared for the colony formation assay. Cells were plated in 6-well plates in media containing 10% FBS at approximately 200 cells per well. Colony formation was then allowed to proceed for two weeks. Cells were washed with 1 ml of PBS, fixed, stained with 500 μl of 0.1% crystal violet solution for 20 min, and finally washed three times with 1 ml of water. The fixed cell colonies were allowed to air dry. The clone formation rate was calculated.
Cell invasion and migration assay
Huh7 and HCCLM3 cells were infected with ATAD2 siRNA for 48 h. Cells were then seeded onto a synthetic basement membrane present in the inset of a 24-well culture plate. In the invasion assay, polycarbonate filters coated with 50 μL Matrigel (1:9, BD Bioscience) were placed in a Transwell chamber (Costar). In the migration assay, no Matrigel was placed in the chambers. Fetal bovine serum was added to the lower chamber as a chemoattractant. Cells were then incubated at 37°C and allowed to invade through the Matrigel barrier for hours. After incubation, filters were fixed and stained with 0.1% crystal violet solution. Non-invading cells were removed using a cotton swab, and invading cells on the underside of the filter were counted with an inverted microscope.
Real-time polymerase chain reaction gene array
Forty-eight hours after siRNA knockdown, RNA was extracted from the cells using Trizol (Invitrogen) and cleaned with the RNeasy_MinEluteTM Cleanup Kit (Qiagen, Valencia, CA). Subsequently, total RNA was reverse transcribed using SuperScript III Reverse Transcriptase (Qiagen), and complementary DNA was amplified by the polymerase chain reaction (PCR) using 2_Super Array PCR master mix (Qiagen). Real-time PCR was then performed on each sample with the Human Tumor Metastasis RT2 ProfilerTM PCR Array (SuperArray Bioscience) in a Thermal Cycler Dice Real Time System (Takara TP800, Japan) according to the manufacturer’s instructions. Data were normalized to GAPDH levels by the 2-ΔΔCt method.
Luciferase assays
293 T cells were plated into 24-well plates at 80% confluence 24 hours before transfection. A mixture of 200 ng pGL3-3′ UTR, 700 ng pGV214-miR-372, and 100 ng Renilla luciferase plasmid were transfected into 293 T cells using Lipofectamine 2000. Firefly and Renilla luciferase activities were calculated using a dual-luciferase reporter system (Promega, Madison, WI).
Statistical analysis
SPSS 17.0 software for Windows was used. The association between ATAD2 expression and HCC patients’ clinicopathological features was evaluated by the χ2 test. The Wilcoxon test was performed to compare data from the densitometry analysis of mRNA and protein expression. The Kaplan-Meier method was used to calculate the patients’ survival by a log-rank test. A Cox repression model was performed for the univariate and multivariate analysis of prognostic variables. All P values reported are two-sided, and P < 0.05 is considered statistically significant.
Discussion
It was demonstrated that ATAD2 is a novel candidate oncogene and possibly a therapeutic target for several types of human cancer [
5,
7,
12,
16]. However, the abnormal expression of ATAD2 and its possible carcinogenesis in HCC have not been studied thus far. IHC showed that ATAD2 expression in the HCC tissues was significantly higher than in adjacent non-tumor tissues. In addition, the upregulation of ATAD2 was observed in the majority of HCCs compared to their adjacent normal liver tissues by Western blotting. These findings provide evidence that the upregulation of ATAD2 may play an important role in HCC tumorigenesis.
Further correlation analyses indicated that the high expression of ATAD2 in the HCC tissues was positively correlated with tumor metastasis. These results demonstrated that the upregulation of ATAD2 in HCC might play an important role in promoting malignant tumors. Similar results were also observed in other human malignancies, such as esophageal, gastric, colon, and breast cancers [
12,
16], in which the overexpression of the ATAD2 gene was often observed in more aggressive tumor subgroups and provided diagnostic value. Furthermore, we found that the high expression of ATAD2 in HCC was a strong and independent predictor of shortened overall survival.
With regard to the gene function, the bromodomain encoded by ATAD2 contributes to the high-affinity recognition of acetylated lysine [
16], which is frequently found in chromatin remodeling and histone acetylation [
17]. The bromodomain may play an important role in cell growth, proliferation, invasion and apoptosis. In the present study, ATAD2 protein was detectable by Western blotting in all seven cell types studied. The Huh7 and HCCLM3 cells had relatively higher levels of endogenous ATAD2, and the depletion of ATAD2 by transfection with siRNA in these two cell lines led to a G1 phase cell cycle arrest and reduced cell growth/proliferation. In addition, siRNA-mediated ATAD2 knockdown could significantly inhibit cell migration and cell invasion. This result was in agreement with previous studies that showed that ATAD2 was closely involved in several key regulatory mechanisms to control cell proliferation or tumor metastasis through its structure domain [
18,
19]. Taken together, these data provide evidence that ATAD2 is not only important in HCC cell proliferation but also involved in carcinoma cell migration and invasion.
However, the molecular mechanisms by which ATAD2 regulates cancer cell proliferation and invasion remain unclear. From the Human Tumor Metastasis RT-PCR array with ATAD2 siRNA-treated cells, we identified APC, ITGB3, FXYD5, ITGA7 and CTNNA1 (α-catenin), which are related to cell adhesion; among these, the expression of APC and CTNNA1 changed by at least 3-fold.
APC, which negatively regulates β-catenin, plays an important role in the development of HCC. In mice lacking APC in the liver, β-catenin was stabilized, which significantly increased the incidence of HCC [
20]. Colnot et al. [
20] found that 67% of APC(lox/lox) mice developed HCC when Cre adenovirus was injected. The β-catenin signaling in these mice was strongly activated, indicating that the loss of APC function led to Wnt signaling pathway activation and caused the occurrence of tumors.
CTNNA1 functions in a complex with CTNNB1, where it acts to tether the cytoplasmic domain of E-cadherin to the cytoskeleton [
21,
22]. Vasioukhin et al. found that inhibiting CTNNA1 destablizes adherens junctions, which weakens the interaction between cells and makes them less sensitive to contact-mediated inhibition [
23]. Previous studies showed that reduced CTNNA1 expression in some solid tumors may be associated with increased malignant behavior [
24‐
26]. However, some studies showed that CTNNA1 was more highly expressed in poorly differentiated HCC than in well-differentiated HCC [
27,
28].
CTNNA1 and CTNNB1 link cadherins to the cytoskeleton at epithelial cell-cell adherens junction complexes [
29]. As is well known, in addition to the formation of the E-cadherin-catenin complex, CTNNB1 is associated with the Wnt/Wingless signal transduction pathway and binds to APC, EGFR and the c-erbB2 proto-oncogene. Thus, whether ATAD2 could affect APC and CTNNA1 to foster tumorigenesis by activating the WNT pathway is a goal for further research.
In the present study, we detected changes in the expression of two genes—upregulated APC and downregulated CTNNA1—after knocking down ATAD2 in Huh7 cells, and we validated these changes at the protein level by Western blotting. It appears that in our HCC cells, ATAD2 regulates cell migration/invasion via the regulation of APC and/or CTNNA1 expression. Through real time PCR detection in 45 paired tissue samples, we also found that ATAD2 was significantly negatively correlated with APC and CTNNA1 expression.
Moreover, we also studied the upstream mechanism regulating ATAD2 expression in HCC. miR-372 belongs to the miR-371-373 gene cluster that also includes miR-93 and miR-302a [
30]. These microRNAs play an important role in the development of many types of human malignant tumors. One report noted that miR-372 is an oncogene in human glioma; it is more highly expressed in glioma cells than in normal brain cells and is closely related to the prognosis of glioma [
31]. In the gastric cancer cells, Cho et al. found that miR-372 could promote cell proliferation and inhibit apoptosis; these biological roles are caused by acting on the downstream LATS2 gene [
32]. LATS2, which is an important tumor suppressor, plays an important role in inhibiting cell proliferation in gastric cancers. In the present study, we found that miR-372 was negatively correlated with ATAD2. Through the luciferase reporter gene assay, we further confirmed that the ATAD2 mRNA 3′ non-coding region (3′ UTR) has a binding site for miR-372. More work is necessary to understand how miR-372 regulates ATAD2 expression.
Competing interests
The authors declare that they have no competing interest.
Authors’ contributions
HYL, HH, YYW, XJL, SGX and XXM carried out the experimental work, YQY provided data analysis and histopathological analysis, YFL provided tumor samples, clinical information, GW designed the study and participated in writing the paper. All authors read and approved the manuscript.