The Pinx1 Gene Downregulates Telomerase and Inhibits Proliferation of CD133⁺ Cancer Stem Cells Isolated from a Nasopharyngeal Carcinoma Cell Line by Regulating Trfs and Mad1/C-Myc/p53 Pathways

Nasopharyngeal carcinoma (NPC) is one of the most common malignant tumors in South China, Southeast Asia, and North Africa. Its annual incidence in China is 20 per 10 million people [1]. NPC is generally treated with radiotherapy (stages I and II) or induced chemotherapy combined with radiotherapy (stages III and IV). Although its current 5-year survival rate has increased to 70–80% [2, 3], its local recurrence and distant metastasis are still the leading causes for treatment failure and death [3]. Recent studies have shown that cancer stem cells (CSCs) are an important factor in causing tumor resistance to drugs and radiation therapy, and are closely related to tumor recurrence and distant metastasis [4]. Therefore, exploring the biological characteristics of CSCs in order to carry out CSC-targeted therapy has gained importance in recent years.

CSCs are seed cells with self-renewal and indefinite differentiation potentials. They can form tumor cells with different degree of differentiation and are the root cause for continuous growth, recurrence, and metastasis of malignant tumors, thereby making traditional radiotherapy and chemotherapy treatments ineffective. More and more scholars have accepted the theory that tumors originate from CSCs. In addition to the above characteristics, CSCs also express several relatively specific biomarkers, such as CD133 [5], CD44 [6], NaNOG [7], and SOX2 [8], and overexpress anti-apoptotic genes, oncogenes, and other malignant biological markers, including telomerase and its hTERT subunit [9], BcL-2 [10], and survivin [11]. CSCs, tumor-like stem cells, or side population cells have been isolated from a variety of malignant tumors, including NPC [12, 13].

We previously studied the role of Pin2/ telomeric repeat factor (TRF)1-interacting telomerase inhibitor 1 (PinX1) gene in NPC and found that it downregulates telomerase and hTERT, a broad-spectrum, tumor-specific biological marker that is highly expressed in vast majority of malignancies but not in normal cells, and inhibits tumor cell proliferation and metastasis [14-16]. Interaction of PinX1 with TRF1 and TRF2 could downregulate hTERT activity and maintain telomere integrity [17, 18], but regulation of the PinX1 gene by these factors has not been completely understood in nasopharyngeal CSCs. It has been proposed that the proto-oncogene c-Myc is one of the major determinants of hTERT expression as it directly binds the hTERT promoter. In addition, Max dimerization protein 1 (Mad1) suppresses c-Myc-mediated transactivation by competing for the ubiquitous binding partner Max and preventing its binding to c-Myc. Consistent with this idea, a previous study revealed that the Mad1/c-Myc pathway is involved in the downregulation of telomerase activity by the PinX1 gene in gastric cancer cells [19]. Furthermore, a recent study has reported that HPV16 E6 suppresses PinX1 transcription by reducing the levels of p53, which enhangces the activity of telomerase [20]. Therefore, we hypothesized that the PinX1 gene downregulates telomerase and inhibits proliferation of nasopharyngeal CSCs through TRFs and Mad1/c-Myc/p53 pathways.

Although much is known regarding the role of PinX1 in regulating telomerase function, less is known about the role of TRFs and Mad1/c-Myc/p53 pathways in telomerase regulation by PinX1 in nasopharyngeal CSCs. The aim of this study was to investigate the roles of TRFs and Mad1/c-Myc/p53 pathways in telomerase regulation in nasopharyngeal CSCs in order to lay the foundation for basic research on nasopharyngeal CSCs and clinical treatment of NPC.

The CNE2 cell line was obtained from Beijing Concord Cell Resource Center and cultured in DMEM (HyClone, USA) supplemented with 5% fetal calf serum (Gibco, USA) in a humidified incubator under 5% CO₂ at 37 ºC.

CNE2 cells were labeled with CD133 antibody (eBioscience, USA). In brief, cells were collected by centrifugation, resuspended in PBE solution (0.5 mL/1 × 10⁸ cells), and incubated with primary CD133 antibody at 10–20 μg/mL at 4 °C for 30 min. The cells were then washed once with 20-fold of PBE, resuspended in PBE (0.3 mL/1 – 10⁸ cells) and incubated with ultrafine magnetic beads coated with corresponding secondary antibody at 8–15 °C for 10–15 min. The beads were then packed into a pre-washed separation column with 0.5 ml PBE in a magnetic field. After three washes with 0.5 ml PBE under gravity, the column was taken out and placed over a new tube. Cells were eluted by addition of 1–2 mL of PBE on the top of the column and pushing down the solution by force from the top. The eluted cells were washed once with medium.

CD133 has been widely used as the marker for identifying and sorting CSCs. In this study, both CD133 and CD44 were used to label CNE2 cells for flow cytometry. In brief, cells were digested, filtered using 40-μm nylon filter (BD Biosciences, USA), counted, resuspended to 5 × 10⁶ cells/mL, and transferred to FACS tube (BD biosciences, USA; 0.5 mL/tube). After centrifugation, cells were resuspended in 40 μL of FACS sorting buffer and blocked by incubating with 20 μL of FcR block buffer for 10 min at room temperature. The cells were then incubated with fluorescent-labeled antibody for 30 min at 4 °C in dark, washed once with 1 mL of FACS sorting buffer, resuspended in 0.5 mL of FACS sorting buffer and subjected to flow cytometric analysis.

CD133⁺ CSCs and CD133^- cells were seeded in 6-well plates at 70% confluency in DMEM medium containing 10% FBS and incubated at 37 °C in a humidified incubator supplemented with 5% CO₂. On reaching confluency, cells were collected in 1.5-mL EP tubes, washed twice with PBS, and mixed with 1 mL/well Trizol reagent to extract mRNA. The expression of NaNOG, SOX2, hTERT ,TRF1, TRF2, Mad1, c-Myc and p53 were measured using qRT-PCR with the primers listed in Table 1. qRT-PCR was performed using ABI7500 real-time qPCR Thermocycle Instrument (ABI, Vernon, CA, USA) and FastStart Universal SYBR Green Master (Rox) (Roche, Indianapolis, IN, USA) with cDNA as the template and GAPDH as the internal reference. The average threshold cycle (Ct) values from three PCR assays were determined and results were calculated with 2^–ΔΔCt and normalized to GAPDH.

The standard CSC induction medium was prepared by supplementing 20 ng/mL HGF, 20 ng/mL bFGF, 10 ng/mL insulin, and B27 into DMEM/F12 medium. The collected CD133⁺ CSCs were washed twice with solutions containing streptomycin and penicillin, and then cultured in an ultra-low 6-well plates (BD Biosciences, USA) containing 2 mL of standard medium per well at density of 1 × 10⁴ cells/well. The medium was refreshed every 3.5 days. When cell diameter was > 50 μm, cells were digested using Accutane and passaged. After 10 days of culture, the number and size of spheres were recorded.

Proliferation, migration, and invasion abilities of CD133⁺ CSCs were compared with those of CD133^- cells using cell counting kit-8 (CCK-8), transwell assay, and scratch assay, respectively, as reported previously [5].

Eight 4-week-old female nude mice with body weight of 17 g were randomly numbered using earrings. A total amount of 1 × 10⁶ CD133⁺ CSCs and CD133- CSCs were subcutaneously injected into the left side and right side, respectively, of the same nude mice, and tumor size was measured once a week in the feeding environment. At 4 weeks after injection, nude mice were sacrificed and tumor grafts were isolated. The size of tumor grafts was measured using the equation V = (a^2*b)/2, where a is the short side length of the tumor graft and b is long side length of the tumor graft [21]. Differences in volume of tumor graft between CD133⁺ CSC and CD133^- CSC-injected sides were compared. The animals were provided by the Animal Laboratory of the Southern Medical University. All experiments were approved by the Laboratory Animal Committee and conducted in accordance with the National Laboratory Animal Care and Maintenance Guide.

The 987-bp PinX1 cDNA was amplified using primers designed based on the PinX1 gene sequence and cloned into vector pcDNA3.0. All primers were fragments were digested with KpnI and XhoI, gel purified, and cloned into synthesized by Bioengineering Technology Co., Ltd. (Shanghai, China), and their sequences are as follows: PinX1-F1, GGGGTACCATGTCTATGCTGGCTGAACGTCG, PinX1-R1, CCGCTCGAGTCATTTGGAATCTTTCTTCTTCTTCTTTT. In brief, mRNA was reverse transcribed into cDNA and PinX1 was amplified using PCR. The amplified fragment was analyzed by 1% agarose gel electrophoresis at 120 V for 25 min and purified. The purified the vector. The recombinant plasmid was identified and sequenced. The obtained PinX1 plasmid as well as the control vector were transfected into CD133+ CSCs and the expression of PinX1 was detected by western blotting and qRT-PCR.

Three siRNA sequences (Table 2) were synthesized and their silencing efficiency was detected by qRT-PCR and western blot. The siRNA with best silencing efficiency was used in follow-up experiments.

In vitro sphere formation, proliferation, migration, invasion and apoptosis of CD133⁺ CSCs after PinX1 overexpression and silencing were compared with those of control using the methods described above.

The effects of PinX1 overexpression and silencing on expression of regulatory factors TRF1, TRF2, hTERT, Mad1, c-Myc and P-53 were measured using qRT-PCR and western blotting. Western blot analysis was performed as previously described [22]. The following primary antibodies were used: TRF1, TRF2, hTERT, Mad1, c-Myc and P-53 (1: 500, Cell Signaling Technology, Boston, USA ) and β-actin (1: 30, 000; Sigma, New York, USA). All horseradish peroxidase-conjugated secondary antibodies were purchased from DAKO (1: 20, 000). The density of the western blot signals was measured with Image J software.

Data were statistically analyzed using SPSS 19.0 software. Differences between two groups were compared using Student’s t-test. Differences among multiple groups were compared using one-way ANOVA test. Differences between paired data within one group were compared using Least Significant Difference (LSD) method. Data were expressed as mean ± SD. P < 0.05 was considered as statistically significant.

The fluorescent labeled CD133 antibody and CD44 antibody were directly incubated with CNE2 cell line for flow cytometric analysis (Fig. 1). The results showed that among the isolated cells, CD133⁺ CSCs accounted for 62.32%, of which, 62.31% were CD133⁺CD44⁺ CSCs and 0.01% were CD133⁺CD44^- CSCs. Three days after culture of the isolated CD133⁺ CSCs, CD133⁺ cells accounted for 63.66% and CD133^- cells accounted for 30.77%. Of the former, 63.65% were CD133⁺CD44⁺ and 0.01% were CD133⁺CD44^-. Of the latter, 30.69% were CD133^-CD44^- and 0.08% were CD133^-CD44⁺ cells. Among the isolated CD133^-cells, CD133⁺ cells accounted for 23.37%, of which 22.35% were CD133⁺CD44⁺ and 0.02% were CD133⁺CD44^-.

The expression of hTERT, NaNOG, SOX2, and PinX1 in CD133⁺ CSCs and CD133^- cells were measured using qRT-PCR as shown in Fig. 2A. It is clear that mRNA levels of NaNOG and SOX2 in CD133⁺ CSCs are 6.22 ± 0.11-fold and 4.95 ± 0.11-fold higher, respectively, than those in CD133^- cells; CD133⁺ cells have higher telomerase activity (all P < 0.001), but lower PinX1 levels, than CD133^- cells. When CD133⁺ CSCs were cultured in serum-free stem cell induction medium for 14 days, they showed strong in vitro sphere formation ability. As shown in Fig. 2B, several stem cell spheres of different sizes and irregular shape were formed in CD133⁺ CSCs, but not in CD133^- cells. Furthermore, in vitro proliferation of CD133⁺ CSCs and CD133^- cells were measured using CCK-8 (Fig. 2C). The results clearly showed the proliferation ability of CD133⁺ CSCs to be 1.26-fold and 1.33-fold higher than that of CD133^- cells at 48 h and 72 h, respectively (p < 0.05), indicating that CD133⁺ CSCs proliferate faster than CD133^- cells. In addition, the distance between the scratch lines of CD133⁺ CSCs was 83.8% and 79.8% of that of CD133^- cells at 24 h and 48 h, respectively (P < 0.05), and the number of migrated CD133⁺ CSCs were 1.53-fold of that of CD133^- cells (P < 0.01) as shown in Fig. 2D and 2E, thereby indicating that CD133⁺ CSCs have stronger migration ability than CD133^- cells. Moreover, the number of invading CD133⁺ CSCs were 1.56-fold that of CD133^- cells (P < 0.01) (Fig. 2F), indicating that CD133⁺ CSCs are more invasive than CD133^- cells.

The in vivo tumorigenic ability of CD133⁺ CSCs and CD133^- cells in nude mice is shown in Fig. 3. It is evident that after 3 days of injection, a tumor appeared first on the left side of the nude mice, where CD133⁺ CSCs were injected, and grew rapidly, and later appeared on the right side of the nude mice, where CD133^- cells were injected. After 6 days of injection, the tumors on the left side grew faster than tumors on the right side. After 4 weeks of injection, all 8 nude mice had tumors on the left side, while only 5 of them had tumors on the right side. In addition, the volume of tumors on the left was significantly greater than that of tumors on the right side (P < 0.05).

The expression of PinX1 in CD133⁺ CSCs was detected by qRT-PCR 48 h after transfection of the constructed PinX1 overexpression plasmid and the corresponding empty plasmid in CD133+ CSCs. Compared with the cells transfected with empty vector, PinX1 mRNA levels in PinX1-overexpressing cells were higher by 6.97 ± 0.58-fold (P < 0.001). Western blotting also showed similar increase in PinX1 protein levels, indicating that the PinX1 overexpression plasmid was successfully transfected (Fig. 4A).

Three PinX1 siRNAs were designed and synthesized. To verify their silencing efficiency, these PinX1 siRNAs as well as the negative control siRNA were transfected into CD133+ CSCs and PinX1 levels were examined in them using qRT-PCR and western blotting (Fig. 4B). It is clear that PinX1 mRNA levels decreased to 0.21 ± 0.01, 0.41 ± 0.01, and 0.36 ± 0.01 in cells transfected with the three PinX1 siRNAs, respectively, compared to the control siRNA. Western blotting also showed that transfection of these three PinX1 siRNA significantly decreased PinX1 protein levels. These results indicate that the three PinX1 siRNA constructs were transfected successfully. Among these three PinX1 siRNAs, siRNA -1 had the best inhibition effect, therefore it was used in the future experiments.

As shown in Fig. 5A, overexpression of PinX1 decreased the sphere formation ability of CD133⁺ CSCs. The number of PinX1-overexpressing CD133⁺ CSCs increased by 1.90- and 2.17-fold at 48 h and 72 h, respectively, which was significantly lower than that of 1.95-and 2.48-fold for CD133⁺ CSCs transfected with empty vector (P < 0.05) (Fig. 5B), indicating that PinX1 overexpression significantly inhibited the growth of CD133⁺ CSCs. Both the scratching experiment (Fig. 5C) and transwell experiment (Fig. 5D) showed that PinX1 overexpression significantly inhibited the migration of CD133⁺ CSCs (P < 0.01). In addition, PinX1 overexpression significantly inhibited the invasion ability of CD133⁺ CSCs (P < 0.01), which was also shown by the transwell experiment (Fig. 5E). Furthermore, the effect of PinX1 overexpression on apoptosis of CD133⁺ CSCs was characterized using flow cytometry (Fig. 5F). It was observed that the apoptotic cells in the lower chamber comprised 22.7 ± 1.49% and 11.92 ± 2.33% of PinX1 overexpression and control CD133⁺ CSCs, respectively. The former was 1.9-fold of the latter (P < 0.05), indicating that PinX1 overexpression could induce apoptosis of CD133⁺ CSCs.

The expression of regulatory factors hTERT, TRF1, TRF2, p53, Mad1, and c-Myc in CD133+ CSCs transfected with empty vector, PinX1 overexpression plasmid, negative control siRNA, and PinX1 siRNA were detected using qRT-PCR (Fig. 6A) and western blot (Fig. 6B). PinX1 overexpression significantly decreases hTERT and c-Myc mRNA levels (all P < 0.001), while it increases TRF1, Mad1, and p53 (all P < 0.001) mRNA levels. On the other hand, PinX1 silencing in CD133+ CSCs significantly decreases TRF1, Mad1, and p53 mRNA levels (all P< 0.001), while it increases hTERT and c-Myc mRNA levels (all P< 0.001).

NPC is a malignant tumor originating in the nasopharyngeal epithelium. At present, radiotherapy or induction chemotherapy plus radiotherapy are the basic modalities for treatment of NPC in clinics [2]. However, local recurrence and distant metastasis are the main causes for treatment failure. In addition, the side effects of radiotherapy and chemotherapy, including acute toxicities such as anemia, leukopenia, thrombocytopenia, mucositis, skin reactions, stomatitis, nausea, vomiting, diarrhea, kidney damage, and treatment-related death, seriously affect its therapeutic effects [23, 24]. Thus, further exploration of factors associated with morbidity and new targets has become the top priority. The discovery of CSCs seems to be a key point to influence the onset and treatment of tumors.

CSCs are a group of “tumor initiating cells” existing in certain tumor tissues or cell lines. They have the characteristics of self-renewal, infinite proliferation, metastasis as well as chemical and radiation resistance, and are closely related to tumor occurrence, development, metastasis, and relapse [4]. Compared with regular tumor cells, CSCs are highly tumorigenic. Very few CSCs could form colonies when cultured in vitro and tumor grafts when injected into experimental animals [25]. Since the first report of isolation of leukemia CSCs in 1997, CSCs have been isolated from breast cancer, glioma, colon cancer, and pancreatic cancer [26-30]. Two research groups [12, 13] have sorted out side population cells (SPCs) from human NPC tissues and cell lines. These cells are label-retaining cells (LRCs) with characteristics of stem cells. Although they only account for 0.3%–2% of the whole population, they can form tumor graft in nude mice and possess very strong tumorigenicity and resistance to radiotherapy and chemotherapy. The authors believed that CSCs would become a new target for effective clinical treatment of NPC. The above findings demonstrated that CSCs may be the main factor affecting malignant tumor occurrence and development, local recurrence, distant metastasis, chemotherapeutic drug tolerance and radiation resistance. Currently, there are a variety of surface markers for sorting and identification of CSCs, such as CD133 and CD44. However, these markers lack high specificity. This leads to the current lack of a well-defined stem cell sorting method and poses a challenge for sorting of CSCs. It has been reported that CD133⁺ cells obtained using CD133 magnetic beads express stem cell-specific marker antigens, NaNOG and SOX2, and have higher abilities for self-renewal, proliferation and differentiation as well as tumorigenesis in nude mice than the CD133^- cells. We believe that CD133⁺ can be used as a specific surface marker of nasopharyngeal CSCs [5]. In addition, CD44⁺ was also used as an additional marker for sorting CSCs [31]. In this study, we used CD133 immunomagnetic beads to isolate CD133⁺ cell population from CNE2 cell line and used CD44 as a marker to co-label CNE2 cells for flow cytometry sorting. In order to enhance the reliability of CSC sorting, the stem cell markers, NaNOG and SOX2, were detected using qRT-PCR in CD133⁺ and CD44⁺ cells. In other words, cells that are CD133⁺, CD44⁺, NaNOG⁺, and SOX2⁺ were used as CSCs in the study. In addition, these cells were further identified as CSCs based on their in vitro sphere-forming and in vivo graft-forming abilities. Moreover, the biological characteristics of the obtained CD133⁺ CSCs were also analyzed. The results showed that CD133⁺ CSCs have higher hTERT activity and stronger proliferation, migration, and invasion abilities than CD133^- cells, indicating that the multi-index sorting and identification method used in our study could be used to reliably obtain nasopharyngeal CD133⁺ CSCs. Although this sorting method is complex and cumbersome, it is more reliable than using a single marker for isolating CSCs. Our study showed that CD44 expression is not significantly different, suggesting that the use of CD44⁺ alone as a marker for CSCs lacks specificity.

Human telomerase is composed of telomerase reverse transcriptase hTERT, telomerase RNA component (hTERC), and telomere-associated protein [32]. Telomerase and hTERT are highly expressed in tumor cells and CSCs, making them more specific tumor biomarkers and therapeutic targets [33, 34]. We have previously reported that in 88% of NPC specimens and 100% of NPC HNE1 cell lines, the expressions of both telomerase and hTERT are elevated and closely related. Furthermore, studies have shown that use of telomerase-directed hTERT promoter to control TK suicide gene-specific expression system (hTERT/TK/pGL3) and enhanced targeting vector can kill NPC cells and tumor graft in nude mice, without damaging the normal control cells, liver and kidney of nude mice [35-38]. These results indicate that inhibition of telomerase activity in tumor cells is a key step in killing tumor cells. Therefore, it is important to explore new factors that inhibit telomerase activity in CSCs. In this study, we found that hTERT expression is higher in CD133⁺ CSCs than CD133^- cells, which provides a basis for CSC-targeted therapy using telomerase as the target.

The PinX1 gene is a tumor suppressor gene found recently. It can specifically inhibit telomerase activity and induce apoptosis of tumor cells [14, 16]. Zhou et al [14]. first cloned the endogenous telomerase inhibitor PinX1 gene and studied its functions. They found that PinX1 is a Pin2/TRF1 binding protein located on chromosome 8p22-23, where loss of heterozygosity (LOH) is frequently observed in a series of human tumor cells. PinX1 and its telomerase inhibitory domain (TID) can bind to hTERT and inhibit the activities of hTERT and telomerase, leading to telomere shortening and cell crisis. Our preliminary study on NPC [15] has shown that in vitro transfection of the exogenous PinX1 gene significantly reduces hTERT and telomerase activities, induces apoptosis, and suppresses the migration ability of NPC cells. These results suggest that the PinX1 gene can be a powerful tool for targeting telomerase in order to resist malignant tumor proliferation and migration. However, the ability of PinX1 to downregulate telomerase activity in CSCs has not been previously reported. In this study, we explored its role in inhibiting proliferation of nasopharyngeal CD133⁺ CSCs using telomerase as a target. The results showed that overexpression of PinX1 in NPC can inhibit hTERT activity, decrease migration and invasion, and induce apoptosis of CD133⁺ CSCs, indicating that utilization of the PinX1 gene to target hTERT may be a new method for targeted therapy of CD133⁺ CSCs in NPC.

Because of the complex role of the PinX1 gene, its acting mechanism is still not clear and may involve a series of regulatory factors. Further exploration of its regulatory mechanism is important to fully understand its role. Among many factors, telomeric correlation factors (TRF1 and TRF2) may play important roles in this area [39-41]. Yoo et al [18]. found that PinX1 could maintain telomere integrity by regulating TRF1 stability. PinX1 overexpression played a role in stabilizing TRF1, and depletion of PinX1 using siRNA resulted in TRF1 degradation, ubiquitination and decreased telomere binding. In other words, hTERT participates in positive and negative regulation of the TRF1 steady-state pathway in a PinX1-dependent manner. Campbell et al [42]. also found that hTERT was downregulated, but PinX1 and telomere-associated proteins TEP1, TRF1, TRF2, and tankyrase were elevated in hematopoietic stem cells of chronic myeloid leukemia. Our results showed that overexpression of PinX1 in nasopharyngeal CD133+ CSCs resulted in upregulation of TRF1 and downregulation of hTERT. On the other hand, silencing of PinX1 decreased TRF1 levels, but upregulated hTERT expression. These results suggest that PinX1 may inhibit hTERT activity by upregulating TRF1, and that TRF1 is involved in inhibiting hTERT activity in a PinX1-dependent manner. However, our results did not show any role of TRF2 in this process. Thus, this function may require the formation of a PinX1-TRF1-hTERT complex.

In addition to TRF1, the products of bHLH (basic helix-loop-helix) domain-containing c-Myc and Mad1 genes may also be involved in the role of the PinX1 gene. They likely bind to E-box (CACGTG) of the hTERT promoter, and regulate hTERT activity at transcriptional level. hTERT is regulated through the Madl/c-Myc pathway, in which c-Myc forms a heterodimer with Max and further binds to the two E boxes on hTERT promoter domain, thereby activating gene transcription. Madl also binds to Max in the same region of hTERT. However, the Madl/Max complex inhibits gene transcription. The change from Myc/Max to Mad1/Max complex switches off gene transcription activated by the former. In other words, Madl and c-Myc function as opposite competitive inhibitors [43-45]. Briatore et al [46]. showed that in patients with myelodysplastic syndromes, hTERT expression is related to transcription factors c-Myc, Mad1 and p53. Mad1 expression was increased, especially in the benign lesions, whereas c-Myc and p53 expressions did not change. The authors suggest that high expression of Mad1 can reduce hTERT expression at an early stage of the disease. Wang et al [19]. constructed PinX1 and PinX1-siRNA expression vectors and transfected them into gastric cancer cells. They found that PinX1- gastric cancer cells showed higher telomerase activity than PinX1⁺ gastric cancer cells, and transfection of PinX1 into PinX1^- gastric cancer cells could inhibit telomerase activity in PinX1^- cells, upregulate Mad1 expression and downregulate c-Myc expression. In contrast, silencing of PinX1 decreases Mad1 expression and upregulates c-Myc expression. The authors thus suggest that PinX1-mediated decrease in telomerase activity may involve the Mad1/c-Myc pathway. In the present study, PinX1 overexpression in nasopharyngeal CD133⁺ CSCs upregulated Madl expression and downregulated c-Myc expression. On the contrary, silencing PinX1 in nasopharyngeal CD133⁺ CSCs downregulated Madl expression and upregulated c-Myc expression. These results suggest that PinX1 may inhibit CD133⁺ CSCs through Madl/c-Myc pathway-mediated inhibition of hTERT activity. Tumor suppressor gene p53 may also be involved in the role of PinX1. A previous study [20] showed that p53 functions as a transcription factor and directly activates PinX1 expression in cervical cancer cells. In addition, inhibition of p53 transcriptional activity inhibited PinX1 expression, leading to increased telomerase activity. PinX1 is a novel target gene of p53, and the p53/PinX1 pathway is a new mechanism through which HPV16 E6 enhances telomerase activity in cervical cancer cells. The results of the present study showed that PinX1 overexpression in CD133⁺ CSCs upregulates p53 levels, while PinX1 silencing downregulates p53 levels, thus suggesting that p53 may play a role in the inhibition of CD133⁺ CSCs through the regulation of PinX1-mediated telomerase activity.

The overexpression of PinX1 in CD133+ CSCs significantly decreased hTERT (P < 0.001), inhibited proliferation, migration, and invasion, induced apoptosis, and significantly decreased c-Myc mRNA levels (P < 0.001), while it increased TRF1, Mad1, and p53 mRNA levels (all P < 0.001). On the other hand, PinX1 silencing in CD133+ CSCs significantly decreased TRF1, Mad1, and p53 mRNA levels (all P < 0.01), while it increased hTERT and c-Myc mRNA levels (all P < 0.05). These results indicate that PinX1 downregulates telomerase activity in CD133+ CSCs, inhibits their proliferation, migration and invasion, and induces apoptosis possibly through pathways involving TRF1, Mad1/c-Myc, and p53. In short, the function and regulatory mechanism of the PinX1 gene are far from clear, especially in CSCs, and its role in other CSCs and the involvement of other factors need to be further explored.

This work was supported by the grants from National Natural Science Foundation of China Youth Fund (81402456) and Science and Natural Science Foundation of Guangdong province, China (2015A030313255).

The authors declare that they have no competing interests.

C. Shen and F. Chen contributed equally to this work.