Epigenetic silence of ankyrin-repeat–containing, SH3-domain–containing, and proline-rich-region– containing protein 1 (ASPP1) and ASPP2 genes promotes tumor growth in hepatitis B virus–positive hepatocellular carcinoma

Authors

  • Jian Zhao,

    Corresponding author
    1. International Joint Cancer Institute & Eastern Hospital of Hepatobiliary Surgery, Second Military Medical University, Shanghai, China
    2. School of Pharmacy and Antibody Engineering Center of Ministry of Education, Shanghai Jiao Tong University and Shanghai Key Laboratory for Cell Engineering and Antibody, Shanghai, China
    • International Joint Cancer Institute, Second Military Medical University, 800 Xiang Yin Road, New Building 10-11th Floor, Shanghai 200433, China
    Search for more papers by this author
    • fax: (86)-21-81870710.

    • These authors contributed equally to this work.

  • Guobin Wu,

    1. International Joint Cancer Institute & Eastern Hospital of Hepatobiliary Surgery, Second Military Medical University, Shanghai, China
    2. Guangxi Cancer Hospital, Guangxi Medical University, Guangxi, China
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Fangfang Bu,

    1. International Joint Cancer Institute & Eastern Hospital of Hepatobiliary Surgery, Second Military Medical University, Shanghai, China
    Search for more papers by this author
    • These authors contributed equally to this work.

  • Bin Lu,

    1. International Joint Cancer Institute & Eastern Hospital of Hepatobiliary Surgery, Second Military Medical University, Shanghai, China
    Search for more papers by this author
  • Anmin Liang,

    1. Guangxi Cancer Hospital, Guangxi Medical University, Guangxi, China
    Search for more papers by this author
  • Lei Cao,

    1. International Joint Cancer Institute & Eastern Hospital of Hepatobiliary Surgery, Second Military Medical University, Shanghai, China
    Search for more papers by this author
  • Xin Tong,

    1. International Joint Cancer Institute & Eastern Hospital of Hepatobiliary Surgery, Second Military Medical University, Shanghai, China
    Search for more papers by this author
  • Xin Lu,

    1. Ludwig Institute for Cancer Research, University of Oxford, Nuffield Department of Clinical Medicine, Oxford, UK
    Search for more papers by this author
  • Mengchao Wu,

    1. International Joint Cancer Institute & Eastern Hospital of Hepatobiliary Surgery, Second Military Medical University, Shanghai, China
    Search for more papers by this author
  • Yajun Guo

    Corresponding author
    1. International Joint Cancer Institute & Eastern Hospital of Hepatobiliary Surgery, Second Military Medical University, Shanghai, China
    2. School of Pharmacy and Antibody Engineering Center of Ministry of Education, Shanghai Jiao Tong University and Shanghai Key Laboratory for Cell Engineering and Antibody, Shanghai, China
    • International Joint Cancer Institute, Second Military Medical University, 800 Xiang Yin Road, New Building 10-11th Floor, Shanghai 200433, China
    Search for more papers by this author

  • Potential conflict of interest: Nothing to report.

Abstract

The ankyrin-repeat–containing, SH3-domain–containing, and proline-rich-region–containing protein (ASPP) family of proteins regulates apoptosis through interaction with p53 and its family members. This study evaluated the epigenetic regulation of ASPP1 and ASPP2 in hepatitis B virus (HBV)-positive hepatocellular carcinoma (HCC) and explores the effects of down-regulation of ASPP1 and ASPP2 on the development of HCC. HCC cell lines and tissues from HCC patients were used to examine the expression and methylation of ASPP1 and ASPP2. The expression of ASPP1 and ASPP2 was diminished in HCC cells by epigenetic silence owing to hypermethylation of ASPP1 and ASPP2 promoters. Analyses of 51 paired HCC and surrounding nontumor tissues revealed that methylation of ASPP1 and ASPP2 was associated with the decreased expression of ASPP1 and ASPP2 in tumor tissues and the early development of HCC. Moreover, ASPP2 became methylated upon HBV x protein (HBx) expression. The suppressive effects on tumor growth by ASPP1 and ASPP2 were examined with RNA interference-mediated gene silence. Down-regulation of ASPP1 and ASPP2 promoted the growth of HCC cells in soft agar and in nude mice and decreased the sensitivity of HCC cells to apoptotic stimuli. Conclusion: ASPP1 and ASPP2 genes are frequently down-regulated by DNA methylation in HBV-positive HCC, which may play important roles in the development of HCC. These findings provide new insight into the molecular mechanisms leading to hepatocarcinogenesis and may have potent therapeutic applications. (HEPATOLOGY 2010;51:142–153.)

The ankyrin-repeat–containing, SH3-domain–containing, and proline-rich-region–containing protein (ASPP) family members are newly identified apoptosis regulation proteins, consisting of ASPP1, ASPP2, and iASPP.1, 2 ASPP1 and ASPP2 function as tumor-suppressor genes and specifically enhance the binding and transactivation of p53 on the promoters of proapoptotic genes.1 Subsequent studies further demonstrate that ASPP1 and ASPP2 can also bind to p63 and p73 and function as common activators of p53 family members.3

Abnormal expression of ASPP family members has been found in a variety of human cancers.4, 5 The expression of ASPP1 and ASPP2 is frequently down-regulated in human breast cancers expressing wildtype p53.1 Aberrant expression of ASPP1 or ASPP2 is also found in lung cancer cells6, 7 and leukemia cells.8, 9 Reduced expression of ASPP2 is related to poor clinical outcomes in diffuse large B-cell lymphoma10 and tumor metastasis and poor recurrence-free survival in breast cancer patients.11, 12

Previous studies have suggested that the expression of ASPP1 and ASPP2 could be activated by E2F,13, 14 or inactivated by DNA methylation.9, 15 Hypermethylation of ASPP1 and ASPP2 promoters is found in several tumor cell lines expressing wildtype p53.15 Methylation of ASPP1 is also reported in acute lymphoblastic leukemia (ALL), and is associated with a high relapse rate and poor prognosis.9

Recent reports have emphasized that epigenetic modifications might play crucial roles in the initiation of cancer. Abnormal gene silencing may benefit the expansion of cells in the early aberrant cloning, and “addict” cancer cells to the subsequent genetic and epigenetic alternations that further promote tumor progression.16, 17 Several tumor suppressor genes have been found frequently hypermethylated in hepatocellular carcinoma (HCC). Analyses of the methylation status of 105 tumor suppressor genes show that methylation of tumor suppressor genes is correlated with HCC development and progression.18 DNA methylation, histone modification, and nucleosomal remodeling are energetically linked in methylation-induced gene silence.19, 20 The involvement of HBV x (HBx) protein in the epigenetic regulation during hepatocarcinogenesis was demonstrated previously, which involves the activation of DNA methyltransferases (DNMTs), and the recruitment of DNMTs and methyl-CpG binding proteins to the target gene promoters.21–24

The expression of ASPP1 and ASPP2 in HCC remains unknown. In this study we analyzed the expression of ASPP1 and ASPP2 and their methylation status in human HCC cell lines and HBV-positive HCC tissues. We also characterized the epigenetic regulation of ASPP1 and ASPP2 by HBx, as well as the tumor-suppressive effects of ASPP1 and ASPP2 in HCC.

Abbreviations

ChIP, chromatin immunoprecipitation; DNMT, DNA methyltransferase; HBV, hepatitis B virus; HBx, hepatitis B virus X; HCC, hepatocellular carcinoma; MSP, methyl-specific PCR; RT-PCR, reverse-transcription polymerase chain reaction; shRNA, short-hairpin RNA; siRNA, small interfering RNA; TUNEL, terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick-end labeling.

Materials and Methods

Cells Lines and HBx-Expressing HepG2 Cells.

Normal liver cell HL7702 and HCC cell lines were cultured at 37°C in an atmosphere containing 5% CO2 in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum. The generation of HBx-expressing HepG2 cells (HepG2-X) is included in the Supporting Data.

RNA Extraction and Reverse-Transcription Polymerase Chain Reaction (RT-PCR).

Total RNA was isolated using the RNeasy mini kit (Qiagen, Hilden, Germany), and genomic DNA was removed using the RNase-Free DNase set (Promega, Madison, WI). First-strand complementary DNA (cDNA) was generated using the Reverse Transcription System Kit (Promega). Expression of ASPP1 and ASPP2 messenger RNA (mRNA) was determined by RT-PCR or quantitative RT-PCR (qRT-PCR) using the LightCycler system (Roche Diagnostics, Basel, Switzerland). Actin was used as an endogenous control to normalize the amount of total RNA in each sample. The primer sequences and PCR conditions can be found in the Supporting Data.

DNA Extraction and Methylation-Specific PCR (MS-PCR) Analysis.

Genomic DNA was extracted from 5 × 106 cells or 10 mg tissue using the TIANamp Genomic DNA Purification Kit (Tiangen Biotechnology, Beijing, China). Genomic DNA was treated with sodium bisulfite as described with the Chemicon's CpGenome Fast DNA Modification Kit (Chemicon, Temecula, CA) and subjected to MS-PCR analysis. Primers specific for methylated and unmethylated ASPP1 or ASPP2 gene were as described.9

MS-PCR products were subcloned into pGEM-T Vector (Promega) and transformed into Escherichia coli. Candidate plasmid clones were sequenced by Generay Biotech (Shanghai, P.R. China).

5-Aza-2′-Deoxycytidine (5-Aza-2′dC) and Trichostatin A Treatment.

2 × 105 HCC cells were seeded in 6-well plate and cultured in medium supplemented with 5-Aza-2′dC (Sigma-Aldrich, St. Louis, MO) at the indicated concentrations for 3-5 days. Alternatively, 0.5 μg/mL Trichostatin A (TSA; Sigma-Aldrich) was added to the indicated cells during the last 24 hours of treatment. Cells were then subjected to RNA or genomic DNA extraction as described.

Construction of Lentivirus (LV) for Short Hairpin RNA (shRNA) of ASPP1 and ASPP2.

Three pairs of cDNA oligonucleotides were designed and synthesized to target ASPP1 or ASPP2 mRNA expression, respectively. The design of the shRNAs was assisted by the use of Web-based software provided by InvivoGen (San Diego, CA; http://www.sirnawizard.com/design.php). Blast searches were performed using the National Center for Biotechnology Information expressed sequence tag database to ensure that the shRNA construct only targeted human ASPP1 or ASPP2 expression. The generation of lentiviruses encoding shASPP1 and shASPP2 can be found in the Supporting Data.

HCC cells were infected with concentrated virus at a multiplicity of infection of 20 in the presence of 8 μg/mL polybrene (Sigma-Aldrich). Supernatant was removed after 24 hours and replaced with complete culture medium. Seventy-two hours after infection the expression of ASPP1 and ASPP2 was confirmed by qRT-PCR and western blot.

Western Blot.

Total cell lysate was prepared in 1× SDS buffer. Proteins at the same amount were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto PDVF membranes. After probing with individual antibodies, antigen-antibody complex was visualized by enhanced chemiluminescence reagents Supersignal (Pierce Biotechnology, Milwaukee, WI). The antibodies specific against ASPP1 (LX54.2) and ASPP2 (DX54.10) were as described.1, 13

Chromatin-Immunoprecipitation (ChIP) Analysis.

ChIP analysis was performed using the Chromatin Immunoprecipitation Assay Kit (Upstate Biotechnology, Lake Placid, NY). Antibodies used for ChIP were anti-acetyl-Histone H3 (Upstate), anti-MeCP2, anti-MBD1 and anti-MBD2 (AVIVA Systems Biology, San Diego, CA), anti-DNMT1, and anti-DNMT3A (Santa Cruz Biotechnology, Santa Cruz, CA). The eluted DNA was purified with phenol/chloroform and a fraction was used as PCR template to detect the presence of the ASPP1 or ASPP2 promoter sequences. Primers used for detecting −25 and +59 of ASPP1 promoter were as described.13 Primers used for detecting −68 and +46 of ASPP2 promoter were forward 5′-CAGTCCGGGGCGAAGAAAGAAAAGGC-3′ and reverse 5′-TCCCTCCTCCGCTCCGAAACCAACTAA-3′. The PCR conditions were as follows: 95°C for 1 minute, then 35 cycles of 94°C for 30 seconds, 68°C for 3 minutes, and a final elongation at 68°C for 3 minutes using Advantage-GC Genomic Polymerase Mix (ClonTech, a TAKARA Bio Company, Shiga, Japan). PCR products were electrophoresed in 2% agarose gel.

MTS Assay and Anchorage-Independent Growth Assay.

Cell growth was measured by 3-(4,5-dimethyl-thiazol-2yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) assay (Promega) in 96-well plates (2500 cells per well) following the instructions of the manufacturer. Each experiment was done in triplicate and repeated three times.

For anchorage-independent growth assay, the cells in single-cell suspension were plated in 0.3% agarose over a 0.6% agarose bottom layer at a density of 500 cells per well in 24-well plates and incubated for 14 days. Finally, the cells were stained and the numbers of colonies greater than 100 μm in diameter were counted.

Detection of Apoptosis.

HCC cells were infected with lentivirus encoding shASPP1, shASPP2, or shNon for 72 hours. Both attached and floating cells were harvested and fixed with 70% ethanol for at least 48 hours. Fixed cells were resuspended in 50 μg/mL propidium iodide and 100 μg/mL RNase for 30 minutes and analyzed by flow cytometry (FACscalibur; Becton Dickinson, Franklin Lakes, NJ). Cell cycle parameters were obtained using curve fitting analysis with the ModFit program (Verity Software, Topsham, ME).

HCC cells grew in 6-well plates at 2 × 105/mL and were transfected with 2 μg p53, ASPP1, ASPP2 plasmids, or pcDNA3 vector as indicated with Fugen (R&D Systems, Minneapolis, MN). Twenty-four hours after transfection the cells were subjected to serum starvation for 48 hours. Apoptotic cells were analyzed by the Annexin V-FITC kit (Jingmei Biotech, Shanghai, China).

In situ apoptosis assay was performed with the Fluorescein FragEL DNA Fragmentation Detection Kit (Calbiochem, San Diego, CA). The formalin-fixed paraffin sections were deparaffinized and incubated with terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick-end labeling (TUNEL) reaction mixture. Apoptotic cells carrying DNA labeled with FITC-dUTP were observed under fluorescence microscope (Olympus, Tokyo, Japan).

Animal Studies.

Male Balb/c nude mice at 6 weeks old were purchased from the Shanghai Experimental Animal Center of Chinese Academic of Sciences (Shanghai, P.R. China). HCC-LM3 cells were infected with LV-shASPP1, LV-shASPP2, and LV-shNon at multiplicity of infection (MOI) 20 and 5 × 106 cells were injected subcutaneously into each mouse (n = 6 mice/group). The tumor volume was calculated according to the formula: V = length × width2 × 0.5. Animals were sacrificed and examined at day 30 after cell inoculation.

Patient Samples and Immunohistochemical Staining.

Fifty-one hepatocellular carcinoma tissues and their corresponding nearby nontumorous livers utilized in this study were obtained from Guangxi Cancer Hospital (Nanning, Guangxi, P.R. China) immediately after surgical resection. The expression of ASPP1 and ASPP2 proteins in the specimens was detected by immunohistochemistry assay. Identification of p53 mutation was obtained by gene sequencing from exon 2 to exon 11 by Shanghai DNA BioTechnologies (Shanghai, P.R. China). Details can be found in the Supporting Data.

Statistical Analysis.

The analyses were carried out using SPSS 13.0 for Windows software (Chicago, IL). P-values for dichotomous variables were two-tailed and based on the Pearson chi-square test or the Pearson chi-square test with continuity correction. Continuous variables were analyzed with Student's t test. A value of P <0.05 was considered statistically significant. All recurrence data were updated on September 31, 2006, and all follow-up data were censored at this point.

Results

Methylation of ASPP1 and ASPP2 in HCC Cells.

The expression of ASPP1 and ASPP2 mRNA was examined in seven HCC cell lines and compared with that in normal liver cell line HL7702 by RT-PCR (Fig. 1A). The expression of ASPP1 and ASPP2 was markedly diminished in HCC-97L, PLC/PRF/5, Huh7 cells with mutant p53 gene and smmu7721 cells with wildtype p53, and slightly reduced in HepG2, HCC-LM3 cells with wildtype p53 gene or in Hep3B cells with p53 gene null. To verify that the decreased expression of ASPP1 and ASPP2 in HCC cell lines was due to DNA methylation, HCC cells were treated with DNA-demethylating agent 5-Aza-2′dC. The expression of ASPP1 and ASPP2 was enhanced with the increased amount of 5-Aza-2′dC in Huh7 cells (Fig. 1B), and significantly enhanced in HCC-97L, PLC/PRF/5, and smmu7721 cells (Fig. 1C). The expression of ASPP1 and ASPP2 was further enhanced by the combination of 5-Aza-2′dC and histone deacetylase inhibitor trichostain A, which indicates that histone deacetylation also contributes to the inactivation of ASPP1 and ASPP2 in HCC cells (Fig. 1D).

Figure 1.

ASPP1 and ASPP2 genes are down-regulated by DNA methylation in HCC cell lines. (A) The expression of ASPP1 and ASPP2 mRNA was examined in seven HCC cell lines and compared with that in normal liver cell line HL7702 by RT-PCR. Quantification of ASPP1 and ASPP2 expression by densitometric analysis of the ASPP1 and ASPP2 bands normalized to the actin bands. (B) Huh7 cells were treated with the indicated amount of 5-Aza-2′dC for 3 days; the expression of ASPP1 and ASPP2 mRNA was determined by RT-PCR. (C) HCC-97L, PLC/PRF/5, and smmu7721 cells were treated with or without 50 μM 5-Aza-2′dC for 3 days; the expression of ASPP1 and ASPP2 mRNA was compared by RT-PCR. (D) Huh7 cells were treated with the indicated amount of 5-Aza-2′dC for 3 days together with or without 0.5 μg/mL TSA for another 24 hours; the expression of ASPP1 and ASPP2 was determined by qRT-PCR. Data are shown as the means ± SD from triplicate experiments. *P < 0.05.

We then analyzed CpG islands in ASPP1 (NT_026437) and ASPP2 (NT_004559) promoters using the CPGPLOT program (http://bioweb.pasteur.fr/seqanal/interfaces/cp-gplot.html). The typical CpG islands showing >50% C+G content and an observed/expected (Obs/Exp) CpG frequency of >0.6 were found in ASPP1 gene ranging from −118 to +806 and ASPP2 gene ranging from −510 to +490. MS-PCR was performed to determine the methylation status of ASPP1 and ASPP2 promoters (Fig. 2A). ASPP1 and ASPP2 promoters were unmethylated in normal liver cell HL7702 and in HepG2 cells which had abundant ASPP1 and ASPP2 mRNA expression. In contrast, ASPP1 and ASPP2 were completely methylated in Huh7 cells which had undetectable ASPP1 and ASPP2 mRNA. Partial methylation of ASPP1 and ASPP2 was found in the remaining HCC cells, which had both methylated and unmethylated alleles (Fig. 2B). Treatment with 5-Aza-2′dC decreased the methylated MS-PCR products and increased the unmethylated MS-PCR products in Huh7 cells (Fig. 2C). Methylation of ASPP1 and ASPP2 was further demonstrated by bisulfite sequencing of four clones from MS-PCR products in each cell line. Extensive hypermethylation of ASPP1 and ASPP2 promoters was observed in HCC-97L, PLC/PRF/5, Huh7, and smmu7721 cells, while only a few CpG islands were methylated in HepG2 cells (Fig. 2D). Together, these data demonstrate that hypermethylation of CpG islands results in epigenetic silence of ASPP1 and ASPP2 in HCC cell lines.

Figure 2.

Methylation analysis of ASPP1 and ASPP2 promoters in HCC cell lines. (A) The CpG island regions of ASPP1 and ASPP2 are shown. The translation initiation sites are denoted 1. The regions used in the ChIP assay and MS-PCR are underlined. (B) MS-PCR analyzed the methylation patterns of ASPP1 and ASPP2 promoters in seven HCC cell lines and a normal liver cell line HL7702. (C) MS-PCR analyzed the methylation patterns of ASPP1 and ASPP2 promoters in Huh7 cells after treatment with 5-Aza-2′dC for 3 days. (D) Sequence analysis of the ASPP1 and ASPP2 promoters after bisulfite modification in HCC cells. The black circle indicates the methylated cytosine while the white circle indicates the unmethylated cytosine in the dinucleotide CpG.

Methylation of ASPP1 and ASPP2 Is Associated with the Early Stages of HCC.

To investigate the methylation status of ASPP1 and ASPP2 in HCC specimens, MS-PCR was performed in 51 paired human HCC tissues and their surrounding nontumor tissues from HBV-positive HCC patients. Methylation of ASPP1 and ASPP2 was 11/51 (21.6%) and 18/51 (35.3%) in the tumor tissues, or 8/51 (15.7%) and 12/51 (23.5%) in the surrounding nontumor tissues, respectively (Fig. 3A,B). There was no statistical significance between the tumor tissues and the surrounding tissues (Supporting Table 1). DNA methylation in both tumor and nontumor tissues was only detected in one case for the ASPP1 gene, and three cases for the ASPP2 gene. Only three cases had both ASPP1 and ASPP2 methylation. Altogether, 26/51 (51%) tumors and 17/51 (33.3%) nontumor tissues had ASPP1 and/or ASPP2 methylation. These data demonstrate that hypermethylation of ASPP1 and ASPP2 promoter is a frequent event in HBV-positive HCCs.

Figure 3.

Methylation analysis of ASPP1 and ASPP2 promoters in HCC tissues. Representative results of MS-PCR analysis and sequence analysis of MS-PCR products of ASPP1 (A) and ASPP2 (B) genes in tumor tissues (T) and adjacent nontumor tissues (NT). (C) Representative immunohistochemistry analysis of ASPP1 and ASPP2 expression in HCC tissues (original magnification ×100). (D) The correlations of ASPP1 and ASPP2 methylation with ASPP1 and ASPP2 expression level were analyzed in 50 HCCs. (E) The correlations of ASPP1 and ASPP2 expression levels with p53 gene status were analyzed. (F) The correlations of ASPP1 and ASPP2 methylation with p53 gene status were analyzed. P-values were calculated by Pearson chi-square test or Pearson chi-square test with continuity correction (*). U, unmethylated; M, methylated; L, low expression; H, high expression.

To correlate the expression of ASPP1 and ASPP2 with their methylation status, 50 HCCs were subjected to immunohistochemistry analysis. Low immunostaining of ASPP1 and ASPP2 was found in 21/50 (42%) and 30/50 (60%) cases of tumor tissues, respectively. Representative immunostainings are shown in Fig. 3C. HCCs with low ASPP1 and ASPP2 immunostaining more frequently had DNA methylation than HCCs with high immunostaining (38.1% versus 6.7% in ASPP1, P = 0.018, and 50% versus 15% in ASPP2, P = 0.012, Fig. 3D). These data demonstrate that DNA methylation contributes to the decreased expression of ASPP1 and ASPP2 in HCCs.

The correlations of the expression and the methylation of ASPP1 and ASPP2 with p53 gene status were further analyzed. HCCs harboring the wildtype p53 gene more frequently had decreased ASPP2 expression (P = 0.028, Fig. 3E). No statistical significance was found between down-regulation of ASPP1 and p53 gene status (P = 0.704, Fig. 3E). There was no significant association between methylation of ASPP1 or ASPP2 with p53 gene status as well (P = 0.136 or 0.178, Fig. 3F). However, when both ASPP1 and ASPP2 were counted, HCCs with the wildtype p53 gene more frequently had ASPP1 and/or ASPP2 methylation than HCCs with the mutant p53 gene (63.0% versus 34.7%, P = 0.047, Fig. 3F).

Methylation of ASPP1 and/or ASPP2 in HCCs was not correlated with age, gender, tumor size, tumor stage, or the recurrent time after operation. However, methylation of ASPP1 and/or ASPP2 in the surrounding nontumor tissues was closely related with tumor size (P = 0.031) and tumor stage (P = 0.010, Table 1). This result indicates that methylation of ASPP1 and ASPP2 may participate in the early stage of neoplasia.

Table 1. Associations of ASPP1 and/or ASPP2 Methylation with Clinicopathological Characteristics in HCC Patients
FeatureTumor TissuesP ValueAdjacent TissuesP Value
U (n = 25)M (n = 26)U (n = 34)M (n = 17)
  1. P values are two-tailed and based on the Pearson chi-squared test.

  2. NS, not significant; U, unmethlation; M, methylation.; AJCC, American Joint Committee on Cancer.

Sex  NS  NS
 Male2325 3315 
 Female21 12 
AFP(ng/mL)  NS  NS
 ≤2016 25 
 >202420 3212 
Tumor volume (cm3)  NS  0.031
 ≤51311 1212 
 >51215 225 
Cirrhosis  NS  NS
 Positive2222 3014 
 Negative34 43 
AJCC Stage  NS  0.010
 Stage I-II1211 1112 
 Stage III-IV1315 235 
Recurrence time (months)  NS  NS
 ≤6137 155 
 >61219 1912 

ASPP2 Gene is Down-regulated by HBx Through DNA Methylation.

Since all HCC patients in this study were HBV-positive, we then investigated whether HBV infection contributed to the methylation of ASPP1 and ASPP2 in HCCs. The expression of ASPP2 was significantly decreased at the RNA and protein levels in HepG2 cells stably expressing HBx (HepG2-X), whereas the expression of ASPP1 was only slightly decreased (Fig. 4A). Treatment with 5-Aza-2′dC significantly enhanced ASPP2 expression in HepG2-X cells (Fig. 4B). MS-PCR analysis revealed that the ASPP2 promoter became methylated upon HBx expression (Fig. 4C). To further explore the mechanisms by which HBx selectively regulates ASPP1 and ASPP2 expression, we analyzed DNMT's expression on HBx expression. The expression of DNMT1 and DNMT3A was not enhanced upon HBx expression (Supporting Fig. 1A); however, the binding of DNMT1 and DNMT3A with the ASPP2 promoter, but not the ASPP1 promoter, was greatly enhanced (Fig. 4D). Silence of DNMT3A expression, but not DNMT1, restored ASPP2 expression in HBx-transfected cells (Fig. 4E, Supporting Fig. 1B). ChIP analyses further revealed that expression of HBx enhanced the recruitment of methyl-CpG-binding proteins MeCP2 and MBD1 on ASPP2 promoter, and inhibited the binding of acetylated histone H3 on the ASPP2 promoter (Fig. 4F). These results indicate that ASPP2 is down-regulated by HBx through the recruitment of DNMT1 and DNMT3A on its promoter to initiate DNA methylation, and subsequently increases the binding of methyl-CpG binding proteins on the ASPP2 promoter to suppress ASPP2 expression.

Figure 4.

ASPP2 is methylated by HBx induction. (A) qRT-PCR and western blot analyzed ASPP1 and ASPP2 expression in HepG2 cells stably transfected with HBx or pcDNA3. Data are shown as the means ± SD from triplicate experiments. *P < 0.05. (B) HepG2 cells with HBx or pcDNA3 stable transfection were treated with 20 μM 5-Aza-2′dC for 5 days; the expression of ASPP2 mRNA was detected by qRT-PCR. *P < 0.05. (C) MS-PCR analyzed the methylation patterns of ASPP2 promoter in HepG2 cells with HBx or pcDNA3 stable transfection. (D) ChIP analysis was performed on HepG2-X cell lysates using antibodies against DNMT1, DNMT3A. DNA derived from aliquots of cell lysates are labeled Input. pcDNA3 vector and IgG served as external and internal controls. (E) The expression of ASPP1 and ASPP2 in HepG2 cells stably transfected with HBx or pcDNA3 was detected with western blot after inhibition of DNMT1 and DNMT3A expression by small interfering RNA (siRNA). Quantification of ASPP2 expression by densitometric analysis of the ASPP2 bands normalized to the actin bands. (F) ChIP analysis was performed on HepG2-X cell lysates using antibodies against MeCP2, MBD1, MBD2 and Acetyl-H3. The immunoprecipitated DNA was quantified by densitometric analysis of the bands normalized to the input bands. Data are shown as the means ± SD from triplicate experiments. *P < 0.05.

Promotion of HCC Tumor Growth by Down-regulation of ASPP1 and ASPP2.

To investigate the role of ASPP1 and ASPP2 in the regulation of tumor development, lentiviruses encoding shRNA against ASPP1 or ASPP2 were generated to inhibit ASPP1 or ASPP2 expression. Infection of LV-shASPP1 and LV-shASPP2 reduced the expression of ASPP1 and ASPP2 by about 50% in HepG2 cells compared to LV-shNon infection or mock control, respectively (Fig. 5A).

Figure 5.

Knock-down of ASPP1 and ASPP2 promotes HCC cell growth in vitro and in vivo. (A) HepG2 cells were transfected with lentiviruses encoding shRNA against ASPP1 and ASPP2 for 72 hours; the expression of ASPP1 and ASPP2 was detected by qRT-PCR and western blot. (B-D) HepG2 cells were infected with LV-shASPP1, LV-shASPP2, or LV-shNon for 72 hours, and Huh7 cells were transfected with pASPP1, pASPP2, or pcDNA by M-PEI for 24 hours. Cells were then seeded in 96-well plates for MTS assay (B) or in semisolid soft agar medium to monitor anchorage-independent growth (C,D). Photographs are representative images from each group as indicated under a microscope (left). The bar graphs show the average number of colonies according to colony foci (right). The numbers represent the means of three independent experiments ± SD. *P < 0.05. (E) HCC-LM3 cells infected with LV-shASPP1, LV-shASPP2, or LV-shNon were injected subcutaneously into nude mice. The tumor volume of each group was scored every 5 days. Data represent the means ± SD. **P < 0.01.

Knock-down of ASPP1 or ASPP2 in HepG2 cells and overexpression of ASPP1 or ASPP2 in Huh-7 cells had no obvious effects on cell proliferation as detected by MTS assay (Fig. 5B). However, the anchorage-independent cell growth was significantly enhanced by ASPP1 or ASPP2 silencing, especially in the ASPP2 silencing group. The colony foci greater than 200 μm were found by ASPP1 or ASPP2 silencing, and three colony foci greater than 400 μm were even found in the ASPP2 silencing group (Fig. 5C). In contrast, introduction of ASPP1 or ASPP2 with M-PEI into Huh-7 cells, which could induce gene expression for over 14 days,25 significantly inhibited colony formation. The colony foci greater than 100 μm decreased by about 50% with ASPP1 or ASPP2 overexpression (Fig. 5D).

To further confirm the inhibitory effects of ASPP1 and ASPP2 on tumor growth in vivo, HCC-LM3 cells infected with LV-shASPP1 or LV-shASPP2 were injected into the flank of nude mice. Compared to nonspecific RNA interference, down-regulation of ASPP1 or ASPP2, especially ASPP2, significantly enhanced the growth of HCC-LM3 cells in nude mice. The tumor volume of shASPP1 or shASPP2 modified HCC-LM3 xenografts was 52% or 72% larger than that of shNon-treated xenografts 30 days after implantation (Fig. 5E).

Regulation of Apoptosis by ASPP1 and ASPP2 in HCC Cells.

To investigate the effects of ASPP1 and ASPP2 on apoptosis, ASPP1 and ASPP2 genes were transfected into HCC cells with different p53 status. Serum-starvation caused a 3-fold increase of apoptotic cells in HCC-LM3 cells that had endogenous wildtype p53. Overexpression of p53 did not further enhance apoptosis in HCC-LM3 cells. In contrast, overexpression of ASPP1 and ASPP2 caused 100% and 70% increases of apoptotic cells, respectively (Fig. 6A). This indicates that ASPP1 and ASPP2 could enhance apoptosis in HCC cells harboring the wildtype p53 gene. Interestingly, introduction of ASPP2 but not ASPP1 into Hep3B cells with p53 gene null induced apoptosis to a similar level as p53 did under serum-starvation (Fig. 6B). Introduction of ASPP1 and ASPP2 genes into Huh-7 with p53220Cys induced apoptosis to an extent similar to that of p53 (Fig. 6C).

Figure 6.

Regulation of apoptosis by ASPP1 and ASPP2 in HCC cells. (A-C) HCC-LM3, Hep3B, and Huh-7 cells were transfected with pcDNA or p53, pASPP1, pASPP2 plasmids. Twenty-four hours after transfection, cells were subjected to serum starvation for 48 hours then subjected to Annexin V-FITC/PI double staining. Annexin V-positive and PI-negative staining cells are indicated as apoptotic cells. The expression of ASPP1 and ASPP2 in the three cell lines after transfection was determined by western blot and shown under each bar chart. (D) HCC-LM3 and HepG2 were infected with LV-shASPP1, LV-shASPP2, or LV-shNon for 72 hours. The percentage of apoptotic cells was measured by the accumulation of cells with a sub-G1 DNA content. (E) HepG2 cells were infected with lentiviruses for 72 hours and then treated with 3 mg/mL cisplatin for 24 hours. The percentage of apoptotic cells was measured by the accumulation of cells with a sub-G1 DNA content. Data are shown as the means ± SD from triplicate experiments. *P < 0.05. (F) In situ TUNEL analysis was performed in tumor sections derived from the same mice as in Fig. 5E at day 30 after cell inoculation. The apoptotic nuclei are seen as a green color excited under fluorescence microscopy. Cell nuclei were counterstained with DAPI and seen as blue (magnification ×400). The percentage of apoptotic cells was calculated by counting green-stained nuclei versus blue-stained nuclei from five randomly chosen fields in each section. Data are present as means ± SD. **P < 0.01.

Knock-down of ASPP1 or ASPP2 significantly reduced the apoptotic cells induced by serum starvation in HepG2 or HCC-LM3 cells that had wildtype p53 (Fig. 6D) and attenuated cisplatin-induced apoptosis in HepG2 cells (Fig. 6E). Consistent with the in vitro experimental results, fewer apoptotic cells were found in HCC-LM3 xenografts with shASPP1 or shASPP2 treatment (Fig. 6F). These data indicate that down-regulation of ASPP1 and ASPP2 in HCC may promote tumor progression through inhibition of cell apoptosis.

Discussion

Dysregulation of apoptosis is closely related to the expansion of tumor cells, metastasis, and resistance to chemotherapy.26–28p53 is a key regulator for apoptosis and frequently mutates in various human cancers.29 The proapoptotic function of p53 is closely linked to its antitumor effects. All of the tumor-derived p53 mutants have lost their ability to induce apoptosis. However, only 30% of HCC contains p53 gene mutations. It remains unclear why wildtype p53 fails to suppress tumor growth in the remaining 70% of HCC.

ASPP1 and ASPP2 proteins interact with p53 and its family members, p63 and p73, to promote apoptosis.1, 3 In this study we describe for the first time that ASPP1 and ASPP2 genes are frequently inactivated by hypermethylation in HCC that is HBV-positive. In HCC tumor tissues, ASPP1 and ASPP2 were frequently found methylated, which contributed to the down-regulation of ASPP1 and ASPP2 in HCCs. Importantly, methylation of ASPP1 and ASPP2 in the surrounding nontumor tissues was closely related to the size and the stage of HCCs. A previous study showed that ASPP1 and ASPP2 were frequently down-regulated in breast cancer expressing wildtype p53.1 In this study we found that HCC tumors with the p53 gene wildtype more frequently had ASPP1 and/or ASPP2 gene methylation. Since p53 mutations mostly occur in high-grade tumors, the high frequency of ASPP1 and ASPP2 gene methylation in tumors containing the wildtype p53 gene revealed that methylation of ASPP1 and ASPP2 might contribute to the early development of HCC by attenuating p53-dependent tumor suppression. These results indicate that inactivation of ASPP1 and ASPP2 by hypermethylation is a frequent event in the early development of HCC.

The ASPP2 gene was found more frequently down-regulated and methylated than the ASPP1 gene in HCC tissues. Moreover, HCCs harboring wildtype p53 more frequently had decreased expression of ASPP2. Knock-down of ASPP2 was more effective in promoting the growth of HCC cells in soft-agar and in nude mice. Thus, ASPP2 might play a more important role in the regulation of tumor development in HCC. ASPP2 was first identified as 53BP2, which contains the C-terminus part of ASPP2.30 The importance of ASPP2 in tumor suppression was recently identified in ASPP2-deficient mice.31 ASPP2 heterozygous mice had a 45% tumor incidence over their lifespan, which was three times that in wildtype mice. ASPP2 heterozygous mice also had an increased susceptibility to γ-irradiation-induced tumor development. Besides p53, several proteins have been found to interact with ASPP2, such as Bcl-2, RelA/p65, and hepatitis C virus core protein.32–35 A recent study has found that Drosophila ASPP (dASPP) could interact physically with C-terminal Src kinase (Csk).36 These interactions might contribute to ASPP2-induced cell survival and proliferation. However, the biological significance of these interactions needs to be explored further.

HBx has been found to promote hypermethylation of tumor suppressor genes like IGFBP-3 and E-cadherin by activation of DNMTs, and recruitment of DNMTs and methyl-CpG binding proteins to the promoters.21, 23 Recently, HBx was found to have a direct interaction with DNMT3A to regulate gene expression epigenetically.24 It has been found that the methyl-CpG-binding domain (MBD) protein, MBD1, formed a complex with histone H3-K9 methylase SETDB1 and chromatin assembly factor CAF-1 to regulate ASPP2 expression.37 Here we found that ASPP1 and ASPP2 were differentially regulated by HBx. Overexpression of HBx induced methylation of ASPP2, but not ASPP1. Further analysis revealed that DNMT1 and DNMT3A were recruited to the ASPP2 promoter, but not to the ASPP1 promoter. Thus, the differential regulation of ASPP1 and ASPP2 methylation by HBx might be due to the lacking of DNMTs binding with the ASPP1 promoter. Overexpression of HBx also recruited MeCP2 and MBD1 to the ASPP2 promoter, and released acetylated histone H3 from the ASPP2 promoter. Therefore, HBx might repress ASPP2 expression through regulating the binding of DNMTs and MBD proteins on the ASPP2 promoter.

In this study, we demonstrate that methylation-induced ASPP1 and ASPP2 silence play important roles in the development of HCC, which might serve as potent targets for the development of anti-HCC therapy.

Acknowledgements

The authors thank Dr. Shen Hou and Qirui Liu for technical support.

Ancillary

Advertisement