Low SFRP1 Expression Correlates with Poor Prognosis and Promotes Cell Invasion by Activating the Wnt/β-Catenin Signaling Pathway in NPC

Nasopharyngeal carcinoma (NPC) is one of the leading malignancies of the head and neck; prevalence is low (0.5 cases per 100,000 individuals) in Europe and the United States, but the incidence rate is high (20–50 cases per 100,000 people) in southern China (1). Improvements in intensity-modulated radiotherapy and chemotherapy have improved local control. However, the outcome remains poor for approximately 30% of patients with NPC. Distant metastasis remains the main reason for treatment failure in NPC (2). Despite great efforts over the past decades to better understand the molecular mechanisms underlying migration and invasion in NPC, biomarkers that can accurately identify patients with high risk of distant metastasis remain absent. Therefore, it is urgent to identify new molecular prognostic markers that can predict distant metastasis and guide individualized treatment of patients with NPC.

Metastasis is an important event in cancer development and progression that contributes to the majority of cancer deaths (3). Deregulation of the signaling pathways involved in cell adhesion and migration is likely to lead to tumor metastasis. Currently, several signaling pathways have been identified as being associated with tumor metastasis, one of which is the Wnt/β-catenin signaling pathway (4, 5). The Wnt/β-catenin signaling pathway is associated with many cell biologic processes such as differentiation, proliferation, and migration (6–8). Overactivation of the Wnt/β-catenin signaling pathway contributes to tumorigenesis, proliferation, and migration in several human cancers, such as hepatocellular carcinoma (9), colon cancer (10), and breast cancer (11). Therefore, therapeutic agents that alter abnormally activated Wnt signaling pathway genes provide potential therapeutic targets for cancer therapies (12–14). Secreted frizzled-related protein 1 (SFRP1), a Wnt antagonist on 8p11, plays an important regulatory role in cellular biologic processes (15, 16). It blocks Wnt signaling by hindering Wnt–receptor interactions via the N-terminal cysteine-rich domain homologous to Frizzled proteins. SFRP1 suppresses tumor proliferation and metastasis in several cancers, such as breast (17) and colorectal carcinomas (18). More importantly, SFRP1 has been reported to be associated with cancer therapy. Recently, it has been reported that SFRP1 could suppress NPC cells proliferation in vitro (19). However, it remains unclear whether SFRP1 plays a direct role in NPC metastasis.

Our research was designed to investigate SFRP1 expression levels in NPC tissues and its correlation with the clinical characteristics of a cohort of patients with NPC. Furthermore, we explored the potential roles of SFRP1 in NPC cell migration and invasion to better understand the mechanism of NPC metastasis, which may provide a novel therapeutic target for treating NPC.

We collected 244 formalin-fixed, paraffin-embedded (FFPE) NPC tissues with detailed clinical long-term follow-up data, which had been obtained from 2003 to 2006, from the Sun Yat-sen University Cancer Center (Guangzhou, PR China). The Institutional Ethical Review Board of Sun Yat-sen University Cancer Center approved this study, and written informed consent was obtained from all patients for the use of their biopsy samples. The 7^th edition of the American Joint Committee on Cancer (AJCC) Cancer Staging Manual was used to reclassify the tumor–node–metastasis (TNM) staging. Table 1 lists the clinical features of all patients. No patient had received any antitumor treatments before biopsy sample collection. All patients were treated with conventional 2-dimensional radiotherapy; patients with stage III–IV disease also received platinum-based concurrent chemotherapy. The median follow-up time was 62.97 months (range, 5.2–91.87). The REMARK guidelines (REporting recommendations for tumor MARKer prognostic studies) were followed (20, 21).

Human NPC cell lines (SUNE-1, CNE-1, CNE-2, HNE-1, HONE-1, C666-1) were grown in RPMI-1640 (Invitrogen) supplemented with 10% FBS (Gibco). The human immortalized nasopharyngeal epithelial cell lines (NP69, N₂-Bmi1, and NPEC-tet) were cultured in keratinocyte serum-free medium (Invitrogen) supplemented with bovine pituitary extract (BD Biosciences). All the human immortalized nasopharyngeal epithelial cell lines and the NPC cell lines which had been authenticated were generously provided by Dr. M. Zeng (Sun Yat-sen University Cancer Center). 293FT cells from ATCC were maintained in DMEM (Invitrogen) supplemented with 10% FBS.

Sections obtained from 244 FFPE NPC specimens were deparaffinized, rehydrated, and the endogenous peroxidase activity was blocked. Then, the slides were subjected to citrate-mediated high-temperature antigen retrieval. BSA was used to block nonspecific binding. Subsequently, slides were incubated with anti-SFRP1 antibody (1:100; Abcam) at 4°C overnight and then incubated with biotinylated secondary antibody for 30 minutes at room temperature. A streptavidin–horseradish peroxidase complex and 3,3′-diaminobenzidine were used to detect and visualize the staining. All sections were scored by 2 board-certified pathologists. The staining index was calculated as the product of staining intensity (Score 1: 0, no staining; 1, weak, light yellow; 2, moderate, yellow-brown; 3, strong, brown) and the proportion of positive cells (Score 2: 1, <10%; 2, 10%–35%; 3, 35%–70%; 4, >70%) as previously described (22).

The staining index, with scores of 0, 1, 2, 3, 4, 6, 8, 9, or 12, was used to evaluate SFRP1 expression in NPC tissues. Receiver operating characteristic (ROC) curve analysis was used to select the cutoff value for high and low expression levels (23, 24). The cutoff values were as follows: low SFRP1 expression, staining index score ≤ 4; high SFRP1 expression, staining index score > 4.

Two-step real-time RT-PCR was used to measure the target gene mRNA levels. Total RNA was extracted using TRIzol reagent (Invitrogen). To measure SFRP1 mRNA expression, total RNA was reverse-transcribed with M-MLV reverse transcriptase (Promega) and random primers (Promega) for SFRP1. SYBR Green–based (Platinum SYBR Green qPCR SuperMix-UDG reagents; Invitrogen) quantitative PCR analysis was carried out using a CFX96 Touch sequence detection system (Bio-Rad). Real-time RT-PCR primers for SFRP1 were 5′-ATCTCTGTGCCAGCGAGTTT-3′ (forward, F) and 5′-GGCTTCTTCTTCTTGGGGAC-3′ (reverse, R). GAPDH was used as the endogenous control for SFRP1, and the comparative threshold cycle (2^−ΔΔCT) equation was used to calculate the relative expression level.

The pcDNA3.1-vector, pcDNA3.1-SFRP1, pSin-EF2-puro-vector, and pSin-EF2-puro-SFRP1 plasmids were obtained from Land.Hua Gene Biosciences. All plasmids were verified by DNA sequencing before use. The pcDNA3.1-vector and pSin-EF2-puro-vector plasmids were used as controls.

The pSin-EF2-puro-SFRP1 or pSin-EF2-puro-vector plasmids were transfected into 293FT cells using Lipofectamine 2000 reagent (Invitrogen) to generate stably transfected cell lines. Lentiviral particles were harvested and infected into SUNE-1 cells 48 hours later. After selection with puromycin (Sigma-Aldrich) for 2 weeks, real-time RT-PCR and Western blotting were used to validate the stably transfected cells.

For MTT assay, CNE-2, SUNE-1, and NP69 cells after transfecting with plasmids or siRNAs were seeded at 1,000 cells per well in 96-well plates. The cell viability was measured at 490 nm with a spectrophotometric plate reader at 1, 2, 3, 4, and 5 days. For colony formation assay, cells were plated at 400 cells per well in 6-well plates and cultured for 7 or 12 days. Colonies were counted under the inverted microscope after fixing in 4% paraformaldehyde and staining with 0.5% crystal violet.

CNE-2, SUNE-1, and NP69 cells were transfected with plasmids or siRNAs and grown to near confluence in 6-well plates. After 24-hour serum starving, the monolayers were scratched using a sterile 200-μL tip, followed by washing with serum-free medium to remove the detached cells. Then, the cells were cultured without serum over the next 24 hours. Images of cells migrating at the corresponding wound sites were captured at 0 and 24 hours using an inverted microscope (40×).

Transwell chambers (8-μm pores; Corning) in which the upper surface of the membrane were coated with or without Matrigel (BD Biosciences) before cell seeding were used to test the cell migration or invasive ability. The dishes were placed in a cell culture incubator for 1 hour at 37°C. CNE-2, SUNE-1, and NP69 cells (5 × 10⁴ or 1 × 10⁵) that had been transfected with plasmids or siRNAs and suspended in serum-free medium were plated in the upper chambers. The lower chambers were filled with medium supplemented with 10% FBS. After 12- or 24-hour incubation, cells that had migrated or invaded were fixed, stained, and counted under an inverted microscope (100×).

Forty BALB/c nude mice (4–6 weeks old, male) were purchased from the Medical Experimental Animal Center of Guangdong Province (Guangzhou, China), 5 mice in each group. In total, 5 × 10⁵ CNE-2 cells or 1 × 10⁶ SUNE-1 cells stably overexpressing vector or SFRP1 that had been resuspended in 200 μL PBS were injected into the dorsal flank and the tail veins of the mice. For tumor growth model, tumor size was measured for 2 or 4 weeks. Then, the tumors were dissected and weighted after the mice sacrificing. For lung colonization model, after 6 or 8 weeks, the mice were sacrificed and the lung tissues were harvested, fixed, and paraffin-embedded before 5-μm tissue sections were obtained. All sections were stained with hematoxylin and eosin (H&E) for examination. All animal research was conducted in accordance with the detailed rules approved by the Animal Care and Use Ethnic Committee. All efforts were made to minimize animal suffering.

CNE-2 and SUNE-1 cells were treated with or without 10 μmol/L DAC (Sigma-Aldrich) for 72 hours, with the drug being replaced every 24 hours. An EZ1 DNA Tissue Kit (Qiagen) was used to isolate cellular gDNA. The gDNA (1–2 μg) was treated with sodium bisulfite using an EpiTect Bisulfite kit (Qiagen) according to the manufacturer's instructions. The bisulfite pyrosequencing primers were designed using PyroMark Assay Design Software 2.0 (Qiagen). The primer sequences were as follows: PCR primers: 5′-GTTAAAATTAAGGGTTTTTATTAGGGTAGA-3′ (F); 5′-TCACTCCCAACTCTCCAAAACT-3′ (R); sequencing primer: 5′-TAACAAAAAAACTTCTATTCC-3′. The PyroMark Q96 System (Qiagen) was used for the sequencing reaction and for quantifying methylation levels.

RIPA buffer containing protease inhibitor cocktail (FDbio Science) was used for cell lysis. Total proteins were separated using SDS-PAGE and then transferred to polyvinylidene fluoride membranes (Millipore). The membranes were hybridized with various primary antibodies overnight at 4°C, followed by incubation with species-matched secondary antibodies. Detection was achieved using enhanced chemiluminescence.

The following antibodies were used: SFRP1 (1:2,000; Abcam); β-catenin (1:500; Proteintech); c-Myc (1:2,000; Proteintech); cyclin D1 (1:500; Proteintech); GAPDH (1:5,000; Epitomics); and anti-rabbit IgG antibody (1:5,000; Epitomics). GAPDH was used as the endogenous control.

All statistical analysis was performed using SPSS 16.0 software (SPSS Inc.); differences with P < 0.05 were considered statistically significant. Data were representative of 3 independent experiments and presented as the mean ± SD. The Student t test, Fisher exact test, or the χ² test were used to compare groups. Survival curves were estimated using the Kaplan–Meier method and univariate analysis. Multivariate Cox regression analysis with backward stepwise approach was used to test for independent prognostic factors.

To examine whether the SFRP1 protein expression level is related to the clinical features of patients with NPC, we applied immunohistochemistry to 244 paraffin-embedded NPC tissue samples. Figure 1A–D depicts representative SFRP1 staining in NPC tissue. SFRP1 protein levels were detected in 221 NPC specimens; 119 of 244 (48.8%) patients with NPC had representatively low SFRP1 expression. More patients with low SFRP1 expression died (P = 0.04) and developed distant metastasis (P = 0.018) than those with high SFRP1 expression (Table 1). No statistically significant correlations between SFRP1 expression and patient age, sex, World Health Organization type, viral capsid antigen immunoglobulin A (VCA-IgA), early antigen immunoglobulin A (EA-IgA), T stage, N stage, TNM stage, or locoregional failure were observed.

Kaplan–Meier analysis was used to evaluate overall survival (OS), disease-free survival (DFS), and distant metastasis–free survival (DMFS) in patients with NPC based on the SFRP1 protein expression level. Patients with NPC with low SFRP1 expression had significantly poorer OS [HR, 2.32; 95% confidence interval (CI), 1.36–3.94; P = 0.002], DFS (HR, 1.98; 95% CI, 1.23–3.18; P = 0.005), and DMFS (HR, 2.07; 95% CI, 1.19–3.59; P = 0.009) than those with high SFRP1 expression (Fig. 1E–G). Multivariate Cox regression analyses indicated that patients with low SFRP1 expression and advanced disease had worse OS, DFS, and DMFS than those with high SFRP1 expression, suggesting that both SFRP1 expression and TNM stage are independent prognostic indicators for survival in patients with NPC (Table 2).

We examined SFRP1 mRNA and protein levels in 3 normal nasopharyngeal epithelial cell lines and 6 NPC cell lines using real-time RT-PCR and Western blotting and found that both the SFRP1 mRNA and protein expression levels were significantly downregulated in NPC cell lines (Fig. 2A and B, P < 0.01). MTT and colony formation assays demonstrated that ectopic expression of SFRP1 (Supplementary Fig. S1, P < 0.01) could suppress the cell viability and proliferation in CNE-2 and SUNE-1 cells (Fig. 2C and D, P < 0.05). SFRP1 silence (Supplementary Fig. S2A, P < 0.01) could promote the cell proliferation in NP69 cells (Supplementary Fig. S2B, P < 0.05).

The wound-healing and Transwell assays were used to study the roles of SFRP1 in NPC cell migration and invasion. The wound-healing assay demonstrated that the migration ability of CNE-2 and SUNE-1 cells transfected with SFRP1 was much lower than that in cells transfected with plasmid vector (Fig. 2E). Restoring SFRP1 expression suppressed the migration (Fig. 2F, P < 0.01) and invasive (Fig. 2G, P < 0.01) abilities of NPC cells in an obvious manner. Knocking down SFRP1 expression could promote the migration abilities of NP69 cells (Supplementary Fig. S2C and S2D, P < 0.01).

To determine whether ectopic expression of SFRP1 affected the NPC tumor growth in vivo, we constructed xenograft tumor model by injecting stably overexpressing SFRP1 or vector NPC cells into the dorsal flank of nude mice. The results showed that the tumors grew at a slower rate and had smaller volumes in the stably overexpressing SFRP1 group (Fig. 3A, P < 0.01), and the average tumor weights were also significantly lower in the SFRP1-overexpressing group in both CNE-2 and SUNE-1 cells (Fig. 3B, P < 0.01).

Lung colonization model was used to examine the effect of SFRP1 on NPC progression. CNE-2 and SUNE-1 cells stably overexpressing SFRP1 or vector were constructed for subsequent experiments. Lung colonization models were established by injecting stably overexpressing SFRP1 or vector NPC cells into the tail veins of nude mice. After 6 or 8 weeks of growth, fewer tumor nodes were seen on the lung surfaces of the SFRP1-overexpressing group compared with the control group (Fig. 3C, P < 0.01). In addition, there were significantly smaller and fewer microscopic tumor nodules in the SFRP1-overexpressing group than in vector group (Fig. 3D, P < 0.01).

To explore whether SFRP1 downregulation is associated with promoter hypermethylation in NPC cells, we treated human immortalized nasopharyngeal epithelial cells and NPC cells with or without the demethylation drug 5-aza-2′-deoxycytidine (DAC). Bisulfite pyrosequencing analysis and real-time RT-PCR were used to detect SFRP1 mRNA expression and CpG methylation level changes, respectively. Bisulfite pyrosequencing revealed high methylation levels of the SFRP1 promoter region in CNE-2 and SUNE-1 cells, which were significantly reduced following DAC treatment (Fig. 4A and B, P < 0.01). Compared with cells without DAC treatment, SFRP1 mRNA expression levels were significantly restored in NPC cells while had little changes in human immortalized nasopharyngeal epithelial cells following treatment with DAC (Fig. 4C, P < 0.01).

To explore the mechanism of SFRP1 that acts as a tumor suppressor gene (TSG) in NPC cells, we examined the protein expression profiles of the Wnt/β-catenin signaling genes β-catenin, c-Myc, and cyclin D1 in SFRP1- or vector-transfected CNE-2 and SUNE-1 cells. Cells overexpressing SFRP1 had lower β-catenin, c-Myc, and cyclin D1 protein levels, which indicated that SFRP1 suppressed NPC cell metastasis by inhibiting the Wnt/β-catenin signaling pathway (Fig. 4D).

In this study, we showed that 48.8% of patients with NPC had low SFRP1 expression and that low SFRP1 expression was significantly associated with poor survival. SFRP1 was downregulated in NPC cell lines, and SFRP1 restoration suppressed NPC cell proliferation, migration, and invasion in vitro and lung colonization in vivo. SFRP1 mRNA expression levels in NPC cell lines were significantly restored, and the methylation levels in SUNE-1 and CNE-2 cells were reduced following DAC treatment. Furthermore, ectopic expression of SFRP1 inhibited the Wnt/β-catenin signaling pathway. Our results suggest that SFRP1 acts as a TSG in NPC and can predict the prognosis of patients with NPC.

Unlike other head and neck tumors, NPC has a high rate of local invasion and early distant metastasis (25). The local control rates for NPC have improved significantly following the application of intensity-modulated radiotherapy. However, distant metastasis currently remains the predominant mode of treatment failure. Therefore, it is urgent to better understand the molecular mechanisms of metastasis in NPC to develop novel target treatment methods for patients with NPC.

The Wnt/β-catenin signaling pathway is one of the most important pathways affecting cell biologic processes, which include differentiation, proliferation, and adhesion (4). In normal conditions, signaling begins with the Wnt ligand binding to Frizzled proteins and lipoprotein receptor–related proteins 5 and 6 receptors. Then, β-catenin accumulates and functions as a transcription cofactor with T-cell factor/lymphoid enhancer factor, which regulates the expression of target genes such as c-myc and cyclin D1. Abnormal activation of the Wnt/β-catenin signaling pathway is a common event in several human cancers, including NPC (26–28), and the abnormal methylation status of Wnt antagonists such as Dickkopf proteins, Wnt inhibitory factor 1, and SFRPs may contribute to it (29–32). An SFRP family member, SFRP1, can block Wnt/β-catenin signaling by hindering Wnt–receptor interactions via an N-terminal cysteine-rich domain homologous to Frizzled proteins (33). SFRP1 is hypermethylated and downregulated in various cancers, including breast (34), colorectal (35), and ovarian cancer (36). SFRP1 hypermethylation and downregulation are also associated with poor prognosis in several tumors (34, 37, 38). Moreover, SFRP1 is associated with tumor chemotherapy, and some antitumor drugs inhibit cell growth through the re-expression of SFRP1 (39–41). However, emerging evidence has indicated that SFRP1 may also be highly expressed in carcinomas and promote tumor proliferation or migration, such as in basal-like breast cancer (42), gastric cancer (43), and metastatic renal carcinoma (44). Recently, it has been reported that SFRP1 could suppress NPC cell proliferation in vitro (19). Notably, there is no research on SFRP1 in NPC metastasis. It remains unclear whether SFRP1 plays a direct role in NPC development and progression and whether it can predict distant metastasis and be a potential biomarker to guide individualized treatment of patients with NPC.

Our study demonstrates that SFRP1 is downregulated in NPC cell lines, and 48.8% of clinical specimens showed representative low SFRP1 protein expression levels. Kaplan–Meier analysis showed that patients with NPC with low SFRP1 expression had significantly poorer OS, DFS, and DMFS rates than those with high SFRP1 expression. Multivariate Cox regression analysis indicated that high SFRP1 expression was an independent prognostic indicator for patients with NPC. These results suggest that the SFRP1 expression level can predict the prognosis of patients with NPC and may provide a new therapy target for individualized treatment of patients with NPC.

SFRP1 acts as a suppressor gene in several tumor types by regulating cell proliferation, migration, and invasion. In this research, we used CNE-2 and SUNE-1 cells overexpressing SFRP1 to investigate the function of SFRP1 in NPC. First, we examined the effects of SFRP1 on NPC cells using MTT, colony formation, wound-healing, and Transwell assays, which showed that cell proliferation, invasion, and migration abilities were significantly suppressed in cells overexpressing SFRP1. The role of SFRP1 in NPC proliferation in vitro was consistent with Li and colleagues (19). Nude mice xenografts tumor model and lung colonization model were used to investigate the effects of SFRP1 in vivo and revealed that SFRP1 overexpression inhibited tumor aggressiveness of NPC. Taken together, these results demonstrate that SFRP1 suppresses NPC cell proliferation, migration, and invasion in vitro and lung colonization in vivo.

Epigenetic alterations of TSGs have been reported in the development of several cancers. Hypermethylation of the SFRP1 promoter has been recognized as a common reason for SFRP1 downregulation in tumors. Therefore, we treated NPC cells with the demethylation agent DAC to explore the mechanism of SFRP1 downregulation in NPC. DAC treatment significantly restored SFRP1 mRNA expression levels; bisulfate pyrosequencing analysis demonstrated that the methylation status of SFRP1 was reduced in CNE-2 and SUNE-1 cells after DAC treatment. In summary, SFRP1 was downregulated in NPC due to hypermethylation of the SFRP1 promoter region.

As SFRP1 is a Wnt/β-catenin signaling pathway antagonist, we examined changes to the downstream target genes using Western blotting. β-Catenin, c-Myc, and cyclin D1 were all downregulated after SFRP1 restoration in NPC cells, which suggests that SFRP1 might suppress NPC proliferation and metastasis via the Wnt/β-catenin signaling pathway.

In conclusion, SFRP1 downregulation may result from epigenetic silencing and is associated with worse survival in NPC. Ectopic SFRP1 expression suppresses cell proliferation, migration, and invasion by the blocking Wnt/β-catenin signaling pathway. Our findings also highlight the possibility that SFRP1 is a potential molecular biomarker for predicting the prognosis of patients with NPC and may provide novel therapeutic strategies in NPC.

No potential conflicts of interest were disclosed.

Conception and design: X.-Y. Ren, G.-Q. Zhou, W. Jiang, Y. Sun, N. Liu, J. Ma

Development of methodology: X.-Y. Ren, G.-Q. Zhou, W. Jiang, Q.-M. He

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): X.-Y. Ren, G.-Q. Zhou, Y.-F. Xu, Y.-Q. Li, X.-R. Tang, Q.-M. He, N. Liu, J. Ma

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): X.-Y. Ren, W. Jiang, Y.-Q. Li, X. Wen, X.-J. Yang, N. Liu, J. Ma

Writing, review, and/or revision of the manuscript: X.-Y. Ren, N. Liu, J. Ma

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): X.-Y. Ren, X.-R. Tang, J. Ma

Study supervision: X.-Y. Ren, N. Liu, J. Ma

This work was supported by grants from the National Natural Science Foundation of China (Nos. 81402532, G.-Q. Zhou; 81372409, Y. Sun); the Natural Science Foundation of Guangdong Province (No. S2013010012220, Y. Sun); the Science and Technology Project of Guangzhou City, China (No. 132000507, Y. Sun); the Science and Technology Project of Guangzhou City, China (No. 14570006, J. Ma); the Health & Medical Collaborative Innovation Project of Guangzhou City, China (No. 201400000001, J. Ma); and the National Science & Technology Pillar Program during the Twelfth Five-year Plan Period (No. 2014BAI09B10, J. Ma).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.