microRNA‐124 inhibits stem‐like properties and enhances radiosensitivity in nasopharyngeal carcinoma cells via direct repression of expression of JAMA

Abstract Cancer stem cells (CSCs) are a source of tumour recurrence in patients with nasopharyngeal carcinoma (NPC); however, the function of microRNA‐124 (miR‐124) in NPC CSCs has not been clearly defined. In this study, we investigated the role of miR‐124 in NPC CSCs. qRT‐PCR was performed to measure miR‐124 expression in NPC tissues and cell lines and the effects of miR‐124 on stem‐like properties and radiosensitivity of NPC cells measured. Luciferase reporter assays and rescue experiments were used to investigate the interaction of miR‐124 with the 3′UTR of junctional adhesion molecule A (JAMA). Finally, we examined the effects of miR‐124 in an animal model and clinical samples. Down‐regulation of miR‐124 was detected in cancer tissues and was inversely associated with tumour stage and lymph node metastasis. Overexpression of miR‐124 inhibited stemness properties and enhanced radiosensitivity of NPC cells in vitro and in vivo via targeting JAMA. Up‐regulation of miR‐124 was correlated with superior overall survival of patients with NPC. Our study demonstrates that miR‐124 can inhibit stem‐like properties and enhance radiosensitivity by directly targeting JAMA in NPC. These findings provide novel insights into the molecular mechanisms underlying therapy failure in NPC.

Cancer stem cells (CSCs) constitute a small subpopulation of cancer cells within tumours with stem-like properties. CSCs can self-renew, generate differentiated daughter cells, give rise to heterogeneous tumour tissue and are present in multiple different human malignancies. 3 Wang et al showed that side population (SP) cells sorted from NPC cell lines are a source of CSCs, based on their 20-fold greater ability to form tumours, relative to non-SP cells. 4 CSC markers include aldehyde dehydrogenase 1 (ALDH1), ATP-binding cassette subfamily G member 2 (ABCG2), SRY (sex-determining region Y)-box2 (SOX2) and OCT4, 5 and CSCs can form tumorspheres in serum-free culture, because of their ability to self-renew. 6 CSCs are widely accepted as potential mediators of therapeutic resistance and considered targets for anticancer treatments. [7][8][9] Above all, CSCs are a source of tumour recurrence and resistance to treatment. microRNAs (miRNAs) are small (20-22 nucleotide) non-coding RNA molecules that function as post-transcriptional regulators by negatively modulating the stability or translational efficiency of their target mRNAs. 10 miRNAs have fundamental roles in controlling the fate of CSCs, and dysregulation of miRNAs is implicated in tumorigenesis. 11,12 The stem cell-related miRNA, miR-124, is highly conserved across diverse species and was originally identified as an abundantly expressed in the central nervous system. 13 miR-124 acts as a tumour suppressor in several human cancers, including glioblastoma, as it inhibits tumour cell proliferation and invasion. 14 Moreover, miR-124 plays a crucial role during the transition from neuronal stem cells to neuronal progenitors. 15 Overall, evidence suggests that miR-124 could be important for the maintenance of stem-like properties of cancer cells and neural cells.
Junctional adhesion molecule A (JAMA), also known as JAM-1 and F11R, belongs to a family of cell adhesion molecules that localize to intercellular junctions. JAMA has key roles in tight junction assembly, epithelial barrier function and epithelial-mesenchymal transition (EMT). 16,17 Moreover, JAMA regulates epithelial and endothelial stem-like behaviour 18,19 ; for example, Sugano et al confirmed that JAMA was expressed at high levels in the enriched haematopoietic stem cell (HSC) fraction, whereas multi-lineage colony-forming activity in JAMA-positive cells is higher than that in JAMA-negative cells. 20 Taken together, the evidence strongly supports the function of JAMA in regulating stem-like behaviour of CSCs.
Cancer stem cells can be regulated by aberrant miRNA expression, and miR-124 and JAMA are involved in the development of stem-like traits; however, the role of miR-124 in NPC CSCs has not been clearly defined. In the current study, we investigated the role of miR-124 in NPC CSCs. We found that miR-124 expression was down-regulated in cancer compared with normal tissues and inversely associated with tumour stage and lymph node metastasis.
Overexpression of miR-124 inhibited NPC stem-like properties in vitro and in vivo, whereas down-regulation of miR-124 promoted NPC cell stemness. Furthermore, we showed that miR-124 control stemness through direct regulation of JAMA expression. Specifically, our study demonstrates that miR-124 can inhibit stem-like properties and enhance radiosensitivity by directly targeting JAMA in NPC, suggesting that miR-124 has potential value for diagnosis and as a therapeutic target in NPC.

| Cell culture and reagents
The human NPC cell lines, CNE1, CNE2, HONE1 and SUNE1, were originally purchased from the American Type Culture Collection (ATCC). CNE2 and HONE1 cell lines tested negative for mycoplasma in 2012. The authenticity of cell lines in our study was verified using the DNA finger printing method. 21 Cells were maintained in RPMI 1640 medium supplemented with 10% foetal bovine serum (FBS; HyClone) and antibiotics, according to ATCC cell culture protocols.
The immortalized cell line, NP69, maintained in our laboratory, was a kind gift from Dr Tsao (Cancer Center, Hong Kong University, Hong Kong). NP69 cells were cultured in keratinocyte serum-free medium (K-SFM) (Gibco). HEK293T cells were originally purchased from ATCC and maintained in Dulbecco's modified Eagle's medium (DMEM, Gibco) supplemented with 10% FBS (HyClone). All cells were incubated at 37°C in a humidified atmosphere with 5% CO 2 .
During passaging, tumorspheres were harvested. The number of tumorspheres formed (diameter ≥40 μm) was counted under a microscope after 14 days of culture.

| Flow cytometry analysis for SP
Once cells had reached a logarithmic growth phase, they were analysed by flow cytometry. Cells were digested and washed twice with calcium-and magnesium-free PBS. Cells were then stained with 5 mg/mL Hoechst 33342 (Sigma-Aldrich) in RPMI 1640 medium (supplemented with 2% FBS) at 37°C for 90 minutes in the dark, with mixing at intervals. Subsequently, cells were washed with PBS, and 1 μg/mL propidium iodide was (Sigma-Aldrich) added; cells were kept at 4°C in the dark before sorting by fluorescence-activated cell sorting (BD Biosciences). A subset of the cells was incubated with 50 μmol/L verapamil for 30 minutes at 37°C before adding Hoechst 33342 to determine the region that contained putative SP cells.

| Total RNA extraction and quantitative reversetranscriptase polymerase chain reaction (qRT-PCR)
Total RNA was extracted using TRIzol reagent (Invitrogen), as described in our previous report. 22 For miRNA analysis, 1 μL of stem-loop RT primer (1 μmol/L) and 2 μg of total RNA were used for first-strand cDNA synthesis with an iScript™ cDNA Synthesis Kit (Bio-Rad). For mRNA analysis, total RNA (4 μg) was transcribed into cDNA using the iScript™ cDNA Synthesis Kit (Bio-Rad). Quantitative RT-PCR was performed on an Mx3005P qPCR System (Agilent Technologies) using an iQ™ SYBR ® Green Supermix Kit (Bio-Rad). The thermocycler programme for qPCR was as follows: initial denaturation at 95°C for 10 seconds, followed by 40 cycles of 10 seconds at 95°C and 20 seconds at 55°C. The threshold cycle (Ct value) was recorded. Quantitative RT-PCR assays were repeated at least three times to ensure statistical rigour. Finally, mRNA and miRNA expression levels were calculated from three independent biological replicates. Fold changes in gene expression were calculated by relative quantification using the 2 △△ C T method.

| Caspase-3 activity and colony formation assays
Caspase-3 activity was measured using a caspase-3 Activity Assay Kit (Beyotime Institute of Biotechnology, Nangjing, Jiangsu province, China), as described in our previous report. 23 Briefly, cells transfected with mimic or inhibitor were incubated for 12 hours following irradiation at a dose of 4 Gy. Cells were lysed with lysis buffer (100 μL per 2 × 10 6 cells) for 15 minutes on ice and washed with D-Hank's medium. Cell extracts mixed with Ac-DEVD-pNA substrate were incubated at 37°C for 2 hours. Extracts were analysed by colorimetric measurement of the p-nitroanilide product at 405 nm and results normalized to those of NC assays, to determine the fold change in caspase-3 activity. For colony formation assays, cells transfected with mimic, NC or inhibitor were incubated for 12 hours. Cells exposed to 0, 2, 4, 6 or 8 Gy of 6 MeV X-ray radiation at room temperature were seeded at 500, 1000, 2000, 4000 or 8000 cells, respectively, in 100-mm culture plates. Following irradiation, cells were incubated for 12 days to allow colony formation and surviving fractions (SFs) calculated. Dose-survival curves were fitted based on the single-hit multi-target theory formula, as detailed in our previous report. 23

| Clinical samples and immunohistochemistry
All samples were collected from the Affiliated Cancer Hospital & Institute of Guangzhou Medical University (Guangzhou) and subjected to histopathological examination. A total of 35 NPC and 15 non-cancer nasopharyngitis (NP) tissues specimens were collected for qRT-PCR. For overall survival (OS) analysis, another 121 samples from patients with NPC were collected, as previously described. 24 To ensure that each tissue contained >80% homogeneous cancer cells in its cross-sectional area, haematoxylin-and-eosin-stained sections matched for each NPC tissue were examined. JAMA expression was measured, as previously described, by immunohistochemistry and Western blotting analysis. 17 All patients involved in this study signed an informed consent form. This research was approved by the medical ethics committee of the Affiliated Cancer Hospital & Institute of Guangzhou Medical University (Guangzhou).

| Western blotting
Protein extracts from cells or tumour tissues were mixed with loading buffer, heated at 70°C for 10 minutes, separated on SDS-PAGE gels and transferred to polyvinylidene fluoride (Millipore) membranes.
Membranes were blocked for 2 hours in 5% bovine serum albumin and incubated overnight at 4°C with SP rabbit polyclonal antibodies against

| Statistical analysis
Statistical analyses were performed with SPSS statistical package version 15.0. In vitro experiments were repeated three times and data are presented as the mean ± SE, unless otherwise noted. Statistical differences among groups were assessed using one-way analysis of variance (ANOVA). Mouse xenografts were analysed by the single-hit Poisson model, using the extreme limiting dilution analysis (ELDA) tool (http:// bioinf.wehi.edu.au/softw are/elda). 25 Kaplan-Meier analysis and the log-rank test were used to evaluate differences in OS rates between groups. P-values <.05 were considered statistically significant.

| miR-124 expression is down-regulated in NPC cell lines and tumour tissues, and inversely correlated with clinical stage
To determine the clinical significance of miR-124 expression in NPC, we first analysed its expression in four NPC cell lines and a non-cancerous cell line, NP69. By qRT-PCR, miR-124 was shown to be significantly down-regulated in all four NPC cell lines compared with NP69 cells ( Figure 1A). Moreover, miR-124 expression was examined in 35 human primary NPC clinical tissue and 15 NP biopsy samples and average miR-124 expression was significantly lower in NPC than NP tissues ( Figure 1B). Furthermore, miR-124 expression in patients with NPC was significantly associated with the degree of metastasis to regional lymph nodes and NPC tumour stage of micro-RNA-124 expression in patients with N2 and N3 was lower than in those with N0 and N1 stage disease ( Figure 1C). Additionally, miR-124 expression was correlated with T stage (progression), with miR-124 levels relatively high in T1 and T2 tumours than in T3 and T4 tumours ( Figure 1D). There were no significant correlations between miR-124 expression levels and sex, age or tumour size in patients with NPC (data not shown). These results indicate that reduced miR-124 expression may be important for NPC initiation and progression. were both significantly lower than those for control cells ( Figure 2F).

| miR-124 up-regulation suppresses the stemlike phenotype and enhances radiosensitivity of NPC cells
Hence, caspase-3 activity and clonogenic assays confirmed that miR-124 overexpression rendered NPC cells much more sensitive to irradiation than controls. These results show that miR-124 suppresses NPC stem cell properties in vitro.

| miR-124 down-regulation promotes stemness and reduces NPC cell radiosensitivity
The effect of miR-124 on the stem-like phenotype of NPC cells was further examined. Compared with the control group, significantly F I G U R E 2 miR-124 suppresses stem-like properties and enhances the radiosensitivity of nasopharyngeal carcinoma (NPC) cells in vitro. A, Sorting of side population CNE2 cells using Hoechst 33342. B, Incidence of tumorspheres after 14 days of culture. C, Tumorsphere diameter in CNE2 (left panel) and HONE1 (right panel) cells, presented as frequency distributions (approximately 60 tumorspheres per group were analysed). D, mRNA (upper panel) and protein (lower panel) expression levels in CNE2 cells. E, Caspase-3 activity in NPC cells after irradiation. Bars with different characters indicate significant differences (P < .05). F, Colony-forming ability of CNE2 cells exposed to various doses of irradiation. The survival fraction (SF) assay was analysed using the single-hit multi-target theory formula. Significant differences between the miR-124 mimic and negative control (NC) groups are indicated by *P < .05 and **P < .01 more SP cells were present among NPC cells with miR-124 downregulated ( Figure 3A). Furthermore, results from tumorsphere formation assays indicated that more and larger spheres were formed in cells with miR-124 down-regulated than in controls ( Figure 3B,C).
In addition, we observed significantly increased expression levels of stem cell markers in CNE2 cells with miR-124 down-regulated relative to controls ( Figure 3D).
To examine whether miR-124 increased NPC cell sensitivity to radiotherapy, caspase-3 activity was measured after cell irradiation.
The results indicated that caspase-3 activity was significantly following irradiation in cells with miR-124 down-regulated relative to control cells ( Figure 3E). We then assessed survival and proliferation of NPC cell lines using a clonogenic assay. As expected, SFs of the miR-124 down-regulated population were significantly higher than those of control cells (particularly with doses ≥4 Gy). D 0 and Dq values of miR-124 inhibitor-transfected NPC cells were also both significantly higher than those of control cells ( Figure 3F). Hence, caspase-3 activity and clonogenic assay results support the concept that miR-124 increases NPC cell sensitivity to radiotherapy. Taken together, these results indicate that miR-124 inhibition promotes a stem-like phenotype in NPC cells in vitro.

| Expression of JAMA, a direct target of miR-124, correlates with of NPC progression
To understand the mechanisms underlying the effects of miR-124 on CSCs, we performed in silico prediction of miR-124 target genes. and protein (lower panel) expression levels of CSC markers in CNE2 cells. E, Caspase-3 activity of NPC cells transfected with either a miR-124 inhibitor or NC after irradiation. Bars with different characters indicate significant differences (P < .05). F, Colony-forming ability of NPC cells. The survival fraction for the colony formation assay was analysed. Significant differences between the miR-124 mimic and NC are indicated by *P < .05 and **P < .01 A predicted binding site for miR-124 was detected in the 3'UTR region of JAMA ( Figure 4A). Moreover, JAMA protein expression was reduced in CNE2 cells transfected with miR-124 mimics and significantly increased in cells transfected with an miR-124 inhibitor ( Figure 4B). To determine whether JAMA is a direct target of miR-124, a double-luciferase reporter assay was performed. The results indicated that miR-124 overexpression suppressed luciferase activity from the wild-type JAMA 3'UTR luciferase construct ( Figure 4C), but not from the mutant JAMA 3'UTR. Furthermore, NC miRNA had no influence on luciferase activity from the wild-type JAMA 3'UTR in HEK293T cells.
To further confirm the relationship between miR-124 and JAMA, 35 human primary NPC clinical tissue and 15 NP biopsy samples were evaluated for clinicopathological significance. As shown in Figure 4D, JAMA expression was significantly higher in NPC than in NP samples. Further analysis showed that JAMA expression was lower in stage T1-T2 NPC samples, than T3-T4 stage NPC samples, showing a significant correlation between JAMA levels and tumour stage ( Figure 4E). JAMA expression levels in patients with N2 and N3 were higher than those in patients with N0 and N1 ( Figure 4F).
Moreover, we observed that down-regulation of miR-124 expression correlated with higher JAMA expression in NPC ( Figure 4G). Collectively, these data provide strong evidence that high JAMA expression is closely related to NPC progression, and show that JAMA is an authentic target of miR-124.

| Overexpression of JAMA could partially reverse the stem-like phenotype and reduce radiosensitivity of NPC cells
To elucidate whether the stemness-suppressive effect of miR-124 was mediated by JAMA in NPC cells, we transfected them with miR-124 mimics using a lentivirus vector expressing JAMA lacking its 3′UTR. Flow cytometry results indicated that restoring JAMA significantly increased the percentage of SP cells suppressed by miR-124 ( Figure 5A,B). Furthermore, results from tumorsphere formation assays also indicated that cell lines transfected with miR-124 mimics exhibited a significant decrease in tumorsphere number compared with those transfected with NC miRNA. Conversely, restoring JAMA partially abrogated the inhibitory effect of miR-124 on CSC tumorsphere formation ( Figure 5C). Western blotting data also indicated that stem cell properties were restored by JAMA ( Figure 5D). Subsequently, we evaluated whether caspase-3 activity could be rescued by ectopic expression of JAMA. As expected, increased caspase-3 activity after irradiation, which was induced by miR-124 transfection, was partially abrogated by JAMA overexpression ( Figure 5E,F). Our previous study indicated that JAMA-induced F I G U R E 5 Overexpression of JAMA partially reverses the stem-like phenotype and reduced radiosensitivity of cancer cells. miR-124 mimics were transfected into cells using a lentivirus vector expressing JAMA. JAMA restoration significantly increased SP cells in CNE2 (A) and HONE1 (B) cells suppressed by miR-124. Tumorsphere formation assays indicated that JAMA restoration partially abrogated the inhibitory effect of miR-124 on tumorsphere formation ability by the CNE2 (C) cell line. (D) Expression levels of OCT4, SOX2 and p-Akt were detected by Western blotting. JAMA overexpression blocked increased caspase-3 activity in CNE2 (E) and HONE1 (F) cell lines. After irradiation, JAMA overexpression reduced radiosensitivity of CNE2 (G) and HONE1 (H) cell lines (*P < .05 and **P < .01) EMT may activate the protein kinase B (PKB, also known as Akt) pathway in NPC cells. 17 Thus, expression of key proteins associated with the Akt pathway and stemness was evaluated in xenograft tumours by Western blotting. The results demonstrated that miR-124 caused JAMA down-regulation, and increased levels of phosphorylated Akt and SOX2, indicating that miR-124-induced inhibition of stem-like behaviour may be associated with the Akt pathway.
Moreover, restoration of JAMA partially reduced the radiosensitivity of CNE2 and HONE1 cell lines, as determined by clonogenic assays (Figure 5G,H). Collectively, these results demonstrate that miR-124 targets JAMA directly, resulting in suppression of stem cell properties and enhanced radiosensitivity in NPC cells.

| Up-regulation of miR-124 suppresses NPC cell tumorigenicity in vivo
To confirm the effect of miR-124 on NPC cell stemness, different numbers of CNE2 cells overexpressing miR-124 or control vector were subcutaneously injected into both posterior flanks of nude mice. As shown in Figure 6A and Table 1, the miR-124 and vector groups showed similar tumour formation capability when 1 × 10 6 cells were injected; however, on injection of 1 × 10 4 cells or fewer, the miR-124 group exhibited lower tumorigenicity compared with the vector group. Statistical analysis indicated that the miR-124 group was more able to form tumours than the vector group. These results are consistent with previous reports indicating that CSC and non-CSCs showed similar tumour-forming ability on injection of 1 × 10 6 cells; however, non-CSCs exhibited lower tumorigenicity than CSCs when less than 1 × 10 4 cells were injected. 26,27 In addition, tumours in the miR-124 group grew more slowly than those in the vector group following implantation of either 1 × 10 6 or 1 × 10 4 cells ( Figure 6B). Consistent with tumour size, tumour weights in the miR-124 group were lower than those in the control group ( Figure 6C). The expression levels of key proteins associated with stemness were detected in xenograft tumours using Western blotting. In vivo results demonstrated that miR-124 caused downregulation of JAMA, OCT4 and SOX2, indicating that miR-124 can inhibit stem-like behaviour ( Figure 6D). Based on these results and those of our previous study, we conclude that miR-124 can inhibit stem-like properties and enhance radiosensitivity directly by targeting the 3′UTR of JAMA via the Akt pathway in NPC cells ( Figure 6E). 5-year OS rates were 67.6% and 43.3% in patients with low and high miR-483-5p expression, respectively. In summary, these data suggest that miR-124 is associated with superior prognosis in NPC.

| D ISCUSS I ON
Nasopharyngeal carcinoma is one of the most common malignancies in south-eastern China, 1 and therapy failure in NPC may be due to CSCs. 3 miR-124 can mediate stem cell differentiation via targeting various genes, 28 and further investigation into the role of miR-124 in NPC CSCs is warranted. In this study, we showed that miR-124 expression levels were inversely correlated with tumour stage and lymph node metastasis in patients with NPC. Moreover, miR-124 inhibited NPC stemness both in vitro and in vivo. Mechanistic studies indicated that JAMA was down-regulated by miR-124 through its 3'UTR. These findings provide novel insights into the molecular mechanisms underlying failure to respond to treatment in patients with NPC. miR-124-3p (known as miR-124 or miR-124a) and miR-124-5p (known as miR-124*) are both mature forms of miR-124, a brain-en- via direct binding to the 3'UTR of CDH2, whereas down-regulation of miR-124 was also associated with higher tumour stage and levels of lymph node metastasis. 35 In this study, we demonstrate for the first time that miR-124 is significantly down-regulated in NPC compared with NP, as well as inversely associated with advanced tumour stage and lymph node metastasis. Moreover, miR-124 could suppress the stem-like phenotype and increase sensitivity of NPC cells to radiotherapy, supporting its potential as a tumour suppressor in NPC.
In the present study, we identified JAMA as a novel target of miR-124. Our results suggested that miR-124 directly targeted JAMA by binding to its 3′UTR, resulting in significantly decreased JAMA protein expression. To our best knowledge, these observations provide the first line of evidence that miR-124 acts as a repressor of JAMA, a cell adhesion molecule found on the surface of human platelets, which functions in cell migration and proliferation 18,19 ; however, the role of JAMA in regulating NPC stem-like TA B L E 1 Tumour formation in BALB/c nude mice initiated by CNE2 cell lines transfected with miR-124 or control vectors In conclusion, we found that down-regulation of miR-124 in patients with NPC was inversely associated with advanced tumour stage and positive lymph node metastasis. miRNA-124 inhibits NPC stemlike properties via direct targeting of the JAMA 3′UTR. These findings provide novel insights into the molecular mechanisms underlying failure to respond to treatment in patients with NPC. As such, restoring miR-124 activity represents a promising strategy for NPC therapy.

ACK N OWLED G EM ENTS
Clinical specimens of this study were partly provided by the ENT department. So, we thank the ENT department for their assistance.

CO N FLI C T O F I NTE R E S T S
No conflict of interests exists.

CO N S E NT FO R PU B LI C ATI O N
The author grants the publisher the sole and exclusive licence of the full copyright in the contribution, which license the publisher hereby accepts. The publisher shall have the exclusive right throughout the world to publish and sell the contribution in all languages, without limitation.

PU B LI S H E R ' S N OTE
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

DATA AVA I L A B I L I T Y S TAT E M E N T
Materials and data are available upon request to the corresponding author.