Background
Nasopharyngeal carcinoma (NPC) is a common head and neck cancer arising from the nasopharynx epithelium, with the highest prevalence in southern China [
1,
2]. According to the National Comprehensive Cancer Network (NCCN) guideline, patients with early-stage disease should be treated with radiotherapy alone, while those with advanced-stage disease should receive combined chemoradiotherapy. The tumor-node-metastasis (TNM) staging system is the main determinant for prognostic prediction and treatment choices for NPC patients [
3]. However, the TNM staging system cannot guide the best individualized treatment because of the biological heterogeneity exist among individuals. Approximately 20–30% of NPC patients eventually develop recurrence or distant metastasis [
4]. During the past decades, great efforts have been made to better understand the molecular mechanisms involved in NPC progression. However, biomarkers that can help to accurately select patients with high risk of treatment failure remain absent. Thus, more studies are required to identify novel prognostic biomarkers to guide the individualized treatment for NPC patients.
DNA methylation, a representative epigenetic mechanism, can regulate gene expression through affecting the alternative promoters, retrotransposon elements, and other functional elements without changing the sequence of the nucleotides, which play important roles in the initiation and progression of cancer [
5‐
7]. The dynamic nature of DNA methylation makes it reversible and can serve as an attractive target for cancer treatment [
7,
8]. For the past decades, most studies on aberrantly DNA methylation in cancers focused on CpG islands (CGIs) in the promoter region of genes. CGI acquires hypermethylation to result in gene silencing, whereas DNA hypomethylation is linked to gene reactivation [
5]. CGI hypermethylation has been recognized as one of the key features of cancer [
9]. Recently, growing evidences indicate that there are high frequencies of CGI methylation of tumor suppressor genes in NPC, which contribute to the initiation and progression of NPC [
10‐
13]. However, the roles of numerous aberrant methylation events in NPC are still unclear and warrant further studies.
TIPE family that contains a highly conserved seven α-helixes TIPE2 homology (TH) domain has been recognized as the regulators of inflammation and tumorigenesis. TIPE3 (also known as TNFAIP8L3) is a novel identified TIPE family member with a unique 19 amino acids N-terminal sequence, which has been reported as a transfer protein of lipid second messengers in regulating tumorigenesis [
14]. Based on our previous genome-wide methylation microarray study (GSE52068) [
12], we found that TIPE3 was significantly hypermethylated in NPC tissues in compared with the normal nasopharyngeal epithelial (NPEC) tissues. However, little is known about the effect of CGI hypermethylation on TIPE3 expression and the biological role of TIPE3 in NPC.
In this study, we analyzed the methylation status and mRNA expression levels of TIPE3 across all the solid cancer types in The Cancer Genome Atlas (TCGA) database and our own NPC tissues to identify the effect of TIPE3 CGI methylation on its transcription. Then, the relationships between the TIPE3 CGI methylation levels and clinical features of NPC patients in two large sample sets were analyzed. Furthermore, we investigated the effects of TIPE3 on NPC cell proliferation, migration, and invasion in vitro and in vivo, which may provide a more personalized therapy target for NPC patients.
Methods
Clinical specimens
25 freshly-frozen NPC biopsy samples and 21 normal nasopharyngeal epithelium tissues were collected from Sun Yat-sen University Cancer Center. In addition, a total of 441 formalin-fixed paraffin-embedded NPC biopsy tissue samples with detailed clinical follow-up information were obtained from Sun Yat-sen University Cancer Center (n = 254) and Zhejiang Cancer Hospital (n = 187) between 2004 and 2007. All samples were read and validated by two authoritative pathologists, and hematoxylin and eosin (H&E) staining confirmed that all of the slides contained > 70% tumor cells. All patients were restaged according to the 7th edition of the AJCC Cancer Staging System. No patents received any anti-tumor therapy before biopsy collection. Definitive radiotherapy was applied to all of the patients, and stages III-IV patients also received platinum-based concurrent chemotherapy. Regular clinical assessments were performed and the median follow-up time was 94 months (range: 2–139 months). This research was authorized by the Institutional Ethical Review Boards of both hospitals, and written informed consents were provided by all patients for using their biopsies.
Cell culture and methyltransferase inhibitor treatment
All human immortalized NPEC cells and NPC cell lines were maintained in our own laboratory (Guangzhou, China). NPEC cell lines (NP69, N2-Tert, and N2-Bmi1) were maintained in Keratinocyte serum-free medium (Invitrogen) supplemented with bovine pituitary extract (BD Biosciences). NPC cell lines (CNE1, CNE2, SUNE1, HNE1 and HONE1) were grown in Roswell Park Memorial Institute (RPMI) 1640 (Invitrogen) supplemented with 10% fetal bovine serum (FBS) (Gibco). 293FT cells were maintained in Dulbecco’s modified Eagle’s medium (Invitrogen) supplemented with 10% FBS. For methyltransferase inhibitor treatment, 1.5 × 105 cells were seeded on 60 mm culture dishes. After 24 h, the cells were cultured with or without the methyltransferase inhibitor 5-Aza-2′-deoxycytidine (DAC, 10 μM) for 72 h by replacing the drug every 24 h, and then harvested to extract DNA and RNA.
DNA isolation and bisulfite pyrosequencing
An AllPrep RNA/DNA Mini Kit (Qiagen), QIAamp DNA FFPE Tissue Kit (Qiagen) or EZ1 DNA Tissue Kit (Qiagen) were used to extract the Genomic DNA from fresh-frozen tissues, FFPE tissues, or cell lines, according to the manufacturer’s instructions. An EpiTect Bisulfite Kit (Qiagen) was used to conduct the bisulfite modification of DNA (1-2 μg). The genomic region of TIPE3 chosen for bisulfite pyrosequencing was chosen according to our previous microarray data [
15]. The bisulfite pyrosequencing primers were designed using the PyroMark Assay Design Software 2.0 (Qiagen). The primer sequences for PCR were as follows: 5′-GGGTTTGTAGGT TTATAGTTAATTT-3′ (forward); 5’-CCTCTCCCTAATACTAAACAACAA-3′(reverse); and for sequencing: 5′-TTGTGGGTAAGTGAGGA-3′. The sequencing reaction and methylation level quantification was conducted using the PyroMark Q96 ID System (Qiagen).
RNA extraction and real time RT-PCR
Total RNA from NPC clinical specimens and cell lines was isolated using the TRIzol reagent (Invitrogen). Real-time RT-PCR was performed to test the mRNA expression levels of target gene as previously described [
16]. Briefly, random primers (Promega) and M-MLV reverse transcriptase (Promega) were applied to synthesize the first strand cDNA. SYBR Green-based (Invitrogen) real-time PCR analysis was then performed using the CFX96 Touch™ sequence detection system (Bio-Rad). The primers for TIPE3 amplification were as follows: 5’-GATTGATGACA CCAGCACG-3′ (forward); 5’-TTTGATCGCCACCTTGAT-3′ (reverse). GAPDH was used as an endogenous control, and the comparative threshold cycle (2-ΔΔCT) equation was used to calculate the relative expression levels.
RNA interference, plasmid construction and transfection
The sequence of HA-tagged human TIPE3 (NM_001311175.1) was cloned into plasmid pSin-EF2-puromycin between the EcoR I and Nhe I restriction sites. The pSin-EF2-TIPE3-HA and empty vector, as well as the lentivirus packaging plasmids psPAX2 and pMD2.G, were co-transfected into 293FT cells using the calcium phosphate method, as described previously [
17]. After transfection, lentivirus particles were harvested and used to infect CNE2 and SUNE1 cells. The stably transfected cells were selected using puromycin (Sigma) and confirmed using western blotting. The small interfering RNAs for TIPE3 were purchased from GenePharma (Jiangsu, China), with the following sequences: siTIPE3–1, 5’-GAUGCCACGUUACAAACAATT-3′; siTIPE3–2, 5’-GACUUAAU CAAGGUGGCGATT-3′. CNE1 and HONE1 cells were transfected with siTIPE3s or control (100 nM) using LipofectamineTM RNAiMAX reagent (Invitrogen).
Western blotting
Western blotting was performed to examine the protein levels of target genes, as previously described [
16]. Radioimmunoprecipitation assay (RIPA) buffer containing a protease inhibitor cocktail (FDbio Science) was used for total protein extraction. Total proteins were separated using 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis, and then transferred to the polyvinylidene fluoride membranes (Millipore). After blocking in non-fat milk, the membranes were incubated with mouse monoclonal anti-HA antibody (1:1000; Sigma) at 4 °C overnight, followed by incubation with goat anti-rabbit secondary antibody (1:5000; Sigma). Finally, enhanced chemiluminescence (Thermo) was applied to test the antigen-antibody reaction, and GAPDH was used as a loading control.
For the MTT assay, 1 × 103 cells were seeded per well in 96-well plates. After incubating for the indicated times, 20 μl of MTT (5 mg/mL, BD Biosciences) were added into each well, and then dimethylsulfoxide was used to resolve the crystals. The cell viability was recorded at 490 nm using a spectrophotometric plate reader. For the colony formation assay, 400 cells were plated per well in 6-well plates. After incubating for 7 or 12 days, the colonies were fixed with methyl alcohol, stained with 0.5% crystal violet, and then counted under an inverted microscope.
Transwell migration and invasion assays
Transwell chambers with 8-μm pores in the membrane (Corning), pre-coated without or with Matrigel (BD Biosciences), were used to study the cell migration and invasion abilities. 5 × 104 or 1 × 105 cells suspended in serum-free medium were added into the upper chambers for migration or invasion assays, respectively, while medium supplemented with 10% FBS was placed into the lower chambers. After growing for 12 h (migration assay) or 24 h (invasion assay), the migrated or invaded cells were fixed with methyl alcohol, stained with 0.5% crystal violet, and counted under an inverted microscope (100×).
In vivo xenograft tumor model
Animal experiments were approved by the Animal Care and Use Ethnic Committee of our Center, and all animal handling protocols were conducted based on the detailed principles to minimize animal suffering. 20 BALB/c nude mice (4 weeks old, female) were purchased from the Medical Experimental Animal Center of Guangdong Province (Guangzhou, China). For the xenografted tumor growth model, 1 × 106 SUNE1 cells stably overexpressing TIPE3 or vector were injected into the right and left dorsal flank of mice (n = 5), respectively. After 7 days, the tumor size was measured every 3 days for 3 weeks. Then, the mice were sacrificed and the tumors were dissected and weighted. For the lung colonization model, 1 × 106 SUNE1 cells stably overexpressing TIPE3 or vector were injected into the tail veins of mice (n = 5 in each group). After two months, the mice were sacrificed and their lungs removed. The lung tissues were fixed and paraffin-embedded before cutting into 5-μm slides. One of every ten slides was stained with hematoxylin and eosin (H&E) for microscopic observation.
Statistical analysis
All statistical analyses were performed using Statistical Package for the Social Sciences 19.0 software (SPSS, Chicago, IL, USA), and a p value < 0.05 was considered as statistically significant based on two-sided tests. Data presented as the mean ± SD were calculated from at least three independent experiments. Continuous variables and categorical variables were compared using Student’s t-test, or chi-square and Fisher’s exact tests. The receiver operating characteristic (ROC) curve was used to identify the optimal cut-off value for high and low methylation of TIPE3. The Kaplan-Meier method and univariate Cox regression analysis were applied to estimate survival, and multivariate Cox regression analysis with the backward stepwise method was used to estimate independent prognostic factors. All data in our study have been recorded at Sun Yat-sen University Cancer Center for future reference (RDDB2018000414).
Discussion
Here, we found that the TIPE3 was hypermethylated and the mRNA was downregulated in several human cancers, including NPC. The downregulation of TIPE3 was associated with its CGI hypermethylation. Furthermore, NPC patients with high TIPE3 CGI methylation levels had worse clinical outcomes. Moreover, restoring TIPE3 expression suppressed NPC cell proliferation, migration and invasion in vitro, and inhibited tumor growth and lung metastatic colonization in vivo. Our findings suggested that TIPE3 hypermethylation has a potential to serve as prognostic biomarker and therapeutic target for NPC individualized therapy.
Epigenetic modifications play vital roles in regulating gene expression, such as DNA methylation, histone modifications, nucleosome positioning, and non-coding RNAs [
5]. DNA methylation is among the best characterized epigenetic alterations, which usually occurs at CpG dinucleotides, where DNA methyltransferases catalyze the transfer of a methyl group to cytosine C-5 position to generate 5-methylcytosine. In mammals, genomic CpGs are enriched at CpG-rich short stretches known as CGIs, which are preferentially located at the promoters of genes [
18]. Aberrant DNA methylations have been recognized involving in human cancer causation, progression and therapy [
19,
20]. With the development of powerful technologies, the genome-scale DNA methylation maps of many cancers have been identified nowadays. Global DNA hypomethylation concomitantly with CGI hypermethylation have been identified in many cancers, like acute myeloid leukemia, colorectal cancer, and glioma [
21‐
23]. In this study, through analyzing our previous methylation microarray data, numerous aberrant methylated CpG sites were found in NPC tissues, among which the TIPE3 was found to be remarkably hypermethylated, which was also observed in several human solid cancers in TCGA dataset, such as BRCA, HNSC, and CESC.
Gene expression is a complex process involving the packaging of DNA regulatory regions, chromatin modifying enzymes and transcription factors. Active gene promoters, especially those are CpG-rich and lacking DNA methylation, usually marked by H3K4me3, have extensive lysine acetylation and variant histone H2A.Z to facilitate transcriptional initiation [
5]. Abnormal gains of DNA methylation in promoter CGIs induce gene transcriptional suppression, is a hallmark of human cancers. Here, we identified that TIPE3 expression was downregulated in NPC. The aberrant CGI methylation and expression state could be partially reversed by the demethylation drug in NPC cells. The TIPE3 mRNA level was inversely associated with its CGI methylation level. We also found that high levels of H3K4me3 enrichment, an epigenetic mark of active promoters, overlapping the TIPE3 CGIs in human cell lines from ENCODE. This suggests that this CGI can act as a promoter and further study will be necessary to determine the mechanisms by which it becomes methylated in NPC. Collectively, our results suggested that the downregulation of TIPE3 was related to its CGI hypermethylation, and epigenetic silencing of TIPE3 was a common event in human cancer.
The TIPE3 family consists of four members: TIPE, TIPE1, TIPE2, and TIPE3 [
14,
24]. Recently, the TIPE family members have been recognized as inflammation, immunity, and cancer regulators [
25,
26]. TIPE, the first identified member of this family, can regulate apoptosis and promote tumor metastasis and proliferation [
27,
28]. TIPE1 was reported to be essential for TNF-α-induced cell death [
29]. TIPE2 can maintain immune hemostasis and function as a tumor suppressor [
30,
31]. TIPE3 was found to be upregulated in lung cancer, esophageal cancer, cervical cancer, and colon adenocarcinoma. The unique NT region of TIPE3, which is not seen in other members of the TIPE family, is believed to be responsible for its unique ability to promote cell growth and survival. Furthermore, TIPE3 lacking the NT region appeared to exert a tumor suppression effect [
14]. In the present study, we found that ectopic expression of TIPE3 significantly suppressed NPC cell proliferation and invasion in vitro and in vivo, indicating that TIPE3 might act as a tumor suppressor in NPC and play a dual role in cancer progression. In fact, many genes are reported to act as either tumor suppressors or oncogenes in different cancer types. Nevertheless, the underlying mechanisms for these contradictory roles of TIPE3 in different cancers remain to be determined.
At present, the clinical decision making for NPC patients mainly relies on the TNM staging system [
3], even though it cannot accurately select those at high risk of treatment failure. Over the past decades, numerous studies have focused on developing efficient prognostic molecular biomarkers, such as EBV-DNA, miRNAs, and gene expression [
32‐
34]. However, there is still no effective predictive model for NPC patients. Recently, increasing evidences has demonstrated that both the single-gene loci and the genome-wide profiling indicated a strong potential to predict outcomes in malignant tumors. For single-gene loci, hypermethylation of CDKN2A in colorectal cancer, MGMT in glioblastoma, BRCA1 in breast cancer were reported to be associated with poor clinical outcomes [
35‐
37]. For genome-wide profiling, the methylation gene panel as a prognostic biomarker of prostate, lung, and other cancers have also been identified [
38,
39]. We previously constructed a six-hypermethylated gene panel to predict NPC patients’ survival [
12]. However, the clinical applications of several aberrantly methylated genes in NPC remain unknown. In this study, our findings demonstrated that NPC patients with high TIPE3 CGI methylation level exhibited a significantly shorter OS, DFS, and DMFS compared with patients with low methylation level. These results implied that the TIPE3 CGI methylation level could help to identify a subgroup of patients with high risk of treatment failure and guide more individualized therapy.