Long noncoding RNA PXN‐AS1‐L promotes the malignancy of nasopharyngeal carcinoma cells via upregulation of SAPCD2

Abstract Accumulating evidences highlight the critical roles of long noncoding RNAs (lncRNAs) in a variety of cancers. LncRNA PXN‐AS1‐L was previously shown to exert oncogenic roles in hepatocellular carcinoma. However, the expression, role, and molecular mechanism of PXN‐AS1‐L in nasopharyngeal carcinoma (NPC) malignancy remain unknown. Here, we determined that PXN‐AS1‐L is upregulated in NPC tissues and cell lines. Increased expression of PXN‐AS1‐L predicts worse prognosis of NPC patients. PXN‐AS1‐L overexpression promotes NPC cell proliferation, migration, and invasion in vitro, and NPC tumor growth in vivo. PXN‐AS1‐L silencing suppresses NPC cell proliferation, migration, and invasion in vitro. Mechanistically, PXN‐AS1‐L directly interacts with SAPCD2 mRNA 3′‐untranslated region, prevents the binding of microRNAs‐AGO silencing complex to SAPCD2 mRNA, and upregulates the mRNA and protein level of SAPCD2. SAPCD2 is also increased in NPC tissues. The expression of SAPCD2 is significantly positively associated with that of PXN‐AS1‐L in NPC tissues. Gain‐of‐function and loss‐of‐function experiments demonstrated that SAPCD2 also promotes NPC cell proliferation, migration, and invasion. Furthermore, depletion of SAPCD2 significantly reverses the roles of PXN‐AS1‐L in promoting NPC cell proliferation, migration, and invasion in vitro, and NPC tumor growth in vivo. In conclusion, lncRNA PXN‐AS1‐L is upregulated in NPC and promoted NPC malignancy by upregulating SAPCD2 via direct RNA‐RNA interaction.


| INTRODUCTION
Nasopharyngeal carcinoma (NPC) is one of the predominant head and neck cancers which derived from nasopharyngeal (NP) epithelium. 1 Although radiotherapy with or without neoadjuvant chemotherapy has shown satisfactory treatment results for NPC patients at early stages, most NPC patients at late stages are difficult to treat. 2 Enhancing the understanding of pathogenic mechanisms of NPC is beneficial for the identification of druggable targets for NPC.

| Human tissue specimens
The Medical Ethics Committee of the People's Hospital of Henan Province (Zhengzhou, China) reviewed and approved the use of clinical tissue specimens. A total of 72 fresh NPC tissues and 22 fresh noncancerous NP tissues were acquired at the time of diagnosis with written informed consent from the People's Hospital of Henan Province (Zhengzhou, China). All these specimens were diagnosed by histopathological examination. The performance of this study was in accordance with Declaration of Helsinki.

| Cell culture
Immortalized human normal NP epithelium cell line NP69 and NPC cell lines SUNE1, CNE1, CNE2, HONE1, and HNE1 were acquired from Sun Yat-sen University Cancer Center (Guangzhou, China). NP69 cells were maintained in Keratinocyte/serum-free medium (Invitrogen, Grand Island, NY) supplemented with bovine pituitary extract (BD Biosciences, San Diego, CA). NPC cell lines were maintained in RPMI 1640 medium (Invitrogen) supplemented with 10% fetal bovine serum (Gibco, Grand Island, NY). All these cells were cultured in a humidified incubator containing 5% CO 2 at 37°C.

| Cell proliferation assay
Cell Counting Kit-8 (CCK-8) and Ethynyl deoxyuridine (EdU) incorporation experiments were employed to determine cell proliferation ability. For CCK-8 assay, 3000 cells were seeded per well into 96-well plates. After culturing for indicted time, cell proliferation was evaluated using the Cell Counting Kit-8 (Dojindo Laboratories) in accordance with the instruction. The absorbance values at 450 nm at each time point were collected to plot cell proliferation curves. EdU incorporation experiment was performed with the Cell-Light ™ EdU Apollo ® 643 In Vitro Imaging Kit (RiboBio, Guangzhou, China) in accordance with the instruction. The results were counted using Zeiss AxioPhot Photomicroscope (Carl Zeiss, Oberkochen, Germany) via collecting at least 5 random fields.

| Cell migration and invasion assays
Transwell migration and invasion assays were employed to determine cell migration and invasion ability. Briefly, 50 000 indicated NPC cells re-suspended in 200 μL serum-free medium were seeded into the upper chamber of a transwell insert without (migration) or with (invasion) pre-coated matrigel. Complete medium was added into the bottom wells. After culturing for 48 hours, the cells remain in the upper chamber were removed. The cells migrated or invaded through the chambers were fixed using methyl alcohol, stained using crystal violet, and counted using Zeiss AxioPhot Photomicroscope via collecting at least 5 random fields.

| RNA pull-down
PXN-AS1-L was in vitro transcribed and biotin-labeled from pSPT19-PXN-AS1-L using the Biotin RNA Labeling Mix (Roche) and T7 RNA polymerase (Roche). After being treated with DNase I (Takara) to remove DNA and purified using RNeasy Mini Kit (Qiagen, Shenzhen, China), 3 µg of purified RNA was incubated with 1 mg of whole-cell lysate from SUNE1 cells for 1 hour at 25°C. Next, the complexes were extracted by streptavidin agarose beads (Invitrogen) and the RNA present in the pull-down material was detected by qPCR as described above. Immunoprecipitation Kit (Millipore, Bedford, MA) and an AGO2 specific antibody (Millipore) following the instructions. RIP-derived RNA was detected by qPCR as described above.

| Dual luciferase reporter assay
pmirGLO or pmirGLO-SAPCD2 was co-transfected with pcDNA3.1-PXN-AS1-L or pcDNA3.1 into SUNE1 cells by Lipofectamine 3000 (Invitrogen). After culturing for 48 hours, the Firefly luciferase activity and Renilla luciferase activity were detected by the Dual-Luciferase Reporter Assay System (Promega) in accordance with the instruction.

| Western blot
Protein expression was quantified by western blot. Total proteins were isolated from indicated NPC cells using RIPA lysis buffer (Beyotime, Shanghai, China). Equal amount of proteins was separated using 12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis, followed by being transferred onto nitrocellulose membrane (Millipore). After being blocked using 5% nonfat milk, the membranes were incubated with SAPCD2 (Abcam, Hong Kong, China) or GAPDH (Cell Signaling Technology, Boston, MA) specifically primary antibodies. After 3 washes, the membranes were further incubated with IRdye 700-conjugated goat anti-mouse IgG or IRdye 800-conjugated goat anti-rabbit IgG (Li-Cor, Lincoln, NE). After 3 washes, the membranes were detected on an Odyssey infrared scanner (Li-Cor).

| Xenograft assays
A total of 1 × 10 7 indicted NPC cells re-suspended in 100 μL phosphate buffered saline were subcutaneously injected into the flanks of 4-to 5-week-old female athymic BALB/C nude mice. Subcutaneous tumor volumes were measured every 3 days by a caliper and calculated following the equation "volume = a × b 2 × 0.5 (a, longest diameter; b, shortest diameter)." At the 18th day after injection, the mice were sacrificed and subcutaneous tumors were resected and weighed. The Medical Ethics Committee of the People's Hospital of Henan Province (Zhengzhou, China) reviewed and approved the use of mice. Proliferation marker proliferating cell nuclear antigen (PCNA) immunohistochemistry (IHC) staining was performed on paraffin sections of these subcutaneous tumors with a PCNA primary antibody (Cell Signaling Technology) and a horseradish peroxidase-conjugated IgG (Beyotime). The proteins in situ were visualized using 3,3-diaminobenzidine. Cell apoptosis of subcutaneous tumors was detected by terminal deoxynucleotidyl transferase-mediated dUTP nick end labelling (TUNEL) assay using the In Situ Cell Death Detection Kit (Roche) in accordance with the instruction. The results were collected using Zeiss AxioPhot Photomicroscope and quantified via counting at least 5 random fields.

| Statistical analysis
The GraphPad Prism Software was employed to carry out statistical analyses. For comparisons, Mann-Whitney test, Log-rank test, Pearson chi-square test, one-way ANOVA followed by Dunnett's multiple comparison tests, Student's F I G U R E 1 PXN-AS1-L is upregulated in nasopharyngeal carcinoma (NPC) and associated with poor survival. A, PXN-AS1-L expression levels in 72 NPC tissues and 22 noncancerous nasopharyngeal (NP) tissues were determined by qPCR P < 0.0001 by Mann-Whitney test. B, Kaplan-Meier analysis of the correlation between PXN-AS1-L expression level and overall survival of these 72 NPC patients. The median PXN-AS1-L expression level was used as the cutoff. P = 0.0224 by Log-rank test. C, PXN-AS1-L expression levels in normal NP epithelium cell line NP69 and NPC cell lines SUNE1, CNE1, CNE2, HONE1, and HNE1 were determined by qPCR. Results are displayed as mean ± SD from 3 independent experiments. ***P < 0.001 by one-way ANOVA followed by Dunnett's multiple comparison tests t test, Spearman correlation analysis, or Kruskal-Wallis test was performed as indicated. Significant difference was defined at P < 0.05.

| PXN-AS1-L is upregulated in NPC and correlated with poor survival of NPC patients
To investigate the expression pattern of PXN-AS1-L in NPC, we measured PXN-AS1-L expression in 72 NPC tissues and 22 noncancerous NP tissues by qPCR. As displayed in Figure  1A, PXN-AS1-L is markedly upregulated in NPC tissues compared to NP tissues. Analyzing the correlation between PXN-AS1-L expression levels and clinicopathologic characteristics showed that high expression levels of PXN-AS1-L is positively associated with advanced clinical stages and lymph node metastasis (N classification) in these 72 NPC cases (Table 1). Moreover, survival analysis revealed that NPC patients with higher PXN-AS1-L expression levels have shorter survival time than those of NPC patients with lower PXN-AS1-L expression levels ( Figure 1B). PXN-AS1-L expression levels in normal NP epithelium cell line NP69 and NPC cell lines SUNE1, CNE1, CNE2, HONE1, and HNE1 were measured by qPCR. The results demonstrated that PXN-AS1-L expression levels are elevated in NPC cell lines compared to normal NP epithelium cell line ( Figure  1C). Collectively, these data suggested that PXN-AS1-L is upregulated in NPC and correlated with advanced clinical stage and poor prognosis of NPC patients, which implied that PXN-AS1-L may be involved in the development of NPC.

| The expression of SAPCD2 is positively associated with PXN-AS1-L in NPC tissues
SAPCD2 expression levels in the same 72 NPC tissues and 22 noncancerous NP tissues used in Figure 1A were measured by qPCR. As displayed in Figure 5A, PXN-AS1-L is consistently increased in NPC tissues compared to NP tissues. Moreover, the expression of SAPCD2 is significantly positively correlated with that of PXN-AS1-L in these 72 NPC tissues (r = 0.6329, P < 0.0001) ( Figure 5B).

| SAPCD2 promotes NPC cell proliferation, migration, and invasion
Although SAPCD2 is revealed to function as an oncogene in melanoma, gastric cancer, HCC, and colorectal cancer, the biological roles of SAPCD2 in NPC are still unknown. To determine the biological roles of SAPCD2 in NPC, we constructed SAPCD2 stably overexpressed SUNE1 cells through transfecting SAPCD2 overexpression plasmid. The overexpression efficiency was confirmed using western blot ( Figure   6A). CCK-8 and EdU incorporation experiments both demonstrated that overexpression of SAPCD2 promotes cell proliferation ( Figure 6B,C). Transwell migration experiments demonstrated that overexpression of SAPCD2 promotes cell migration ( Figure 6D). Transwell invasion experiments displayed that overexpression of SAPCD2 promotes cell invasion ( Figure 6E). Furthermore, we constructed SAPCD2 stably silenced SUNE1 cells through transfecting SAPCD2 specific shRNA. The silencing efficiency was confirmed using western blot ( Figure 6F). CCK-8 and EdU incorporation experiments demonstrated that silencing of SAPCD2 suppresses cell proliferation ( Figure 6G,H). Transwell migration experiments displayed that silencing of SAPCD2 suppresses cell migration ( Figure 6I). Transwell invasion assays demonstrated that silencing of SAPCD2 suppresses cell invasion ( Figure 6J). Taken together, these results showed that consistent with PXN-AS1-L, SAPCD2 also promotes NPC cell proliferation, migration, and invasion.

AS1-L in NPC are dependent on the regulation of SAPCD2
To determine whether PXN-AS1-L exerts its oncogenic roles via regulation of SAPCD2, we stably silenced SAPCD2 expression in PXN-AS1-L stably overexpressed SUNE1 cells ( Figure 7A). CCK-8 and EdU incorporation experiments showed that silencing of SAPCD2 attenuated the pro-proliferative roles of PXN-AS1-L overexpression ( Figure 7B,C). Transwell migration assays demonstrated that silencing of SAPCD2 attenuated the pro-migratory roles of PXN-AS1-L overexpression ( Figure 7D).
Transwell invasion experiments showed that silencing of SAPCD2 attenuated the pro-invasive roles of PXN-AS1-L overexpression ( Figure 7E). Furthermore, these constructed SUNE1 cells were subcutaneously injected into nude mice. Tumor volumes were measured every 3 days. Subcutaneous tumors were resected and weighed at the 18th day after injection. As displayed in Figure 7F,G, overexpression of PXN-AS1-L promotes NPC tumor growth in vivo. Silencing of SAPCD2 attenuates the pro-growth roles of PXN-AS1-L overexpression in vivo. Proliferation marker PCNA IHC staining displayed that overexpression of PXN-AS1-L upregulates PCNA expression, which is attenuated by SAPCD2 silencing ( Figure 7H). Apoptosis marker TUNEL staining displayed that overexpression of PXN-AS1-L reduces the number of apoptotic cells, which is reversed by SAPCD2 silencing ( Figure 7I). These data demonstrated that SAPCD2 silencing attenuates both the in vitro and in vivo oncogenic roles of PXN-AS1-L in NPC.

| DISCUSSION
Yuan et al recently reported that splicing factor MBNL3 modulated the alternative splicing of lncRNA PXN-AS1, which generated 2 different isoforms of PXN-AS1. 33 PXN-AS1-L is one of the isoforms which contains the exon 4 and has 863 nucleotides in length, and whereas PXN-AS1-S is another isoform which lacks the exon 4 and has 686 nucleotides in length. 33 They found PXN-AS1-L was upregulated in HCC tissues and had oncogenic roles in HCC. 33 In this study, we focused our attention on lncRNA PXN-AS1-L. Using isoform specific primers, we found that PXN-AS1-L is also increased in NPC tissues and cell lines compared with noncancerous NP tissues and normal NP epithelium cell line, respectively.
Higher expression level of PXN-AS1-L is positively correlated with advanced clinical stage, lymph node metastasis, and poor survival of NPC patients. These data implied that PXN-AS1-L may be a promising prognostic biomarker for NPC. Multicenter studies enrolling more NPC patients can provide stronger evidences for the application of PXN-AS1-L for NPC patients' prognosis. Furthermore, whether PXN-AS1-L is also upregulated in other cancers except HCC and NPC and whether PXN-AS1-L is correlated with outcome of other cancers patients need further exploration. Functional assays revealed that overexpression of PXN-AS1-L promotes NPC cell proliferation, migration, and invasion in vitro. PXN-AS1-L silencing represses NPC cell proliferation, migration, and invasion in vitro. Furthermore, we also found that overexpression of PXN-AS1-L promotes NPC tumor growth in vivo. Therefore, these findings demonstrated that PXN-AS1-L acts as an oncogene in NPC. Our findings also implied that PXN-AS1-L would be a potential therapeutic target for NPC. Previous report has identified the oncogenic roles of PXN-AS1-L in HCC. 33 Thus, we speculate that PXN-AS1-L may be an important oncogene in human cancers. More investigations about the functions of PXN-AS1-L in other cancers can validate this speculation.
The molecular mechanisms exerted by lncRNAs are diverse. Using TCGA dataset, we noted that the expression of PXN-AS1-L is significantly positively associated with SAPCD2 (r = 0.572) in head and neck squamous cell carcinoma. The significant association between PXN-AS1-L expression level and SAPCD2 expression level was further verified in NPC tissues (r = 0.633). Therefore, we further investigated the regulatory effects between PXN-AS1-L and SAPCD2. Our findings revealed that PXN-AS1-L upregulated the mRNA and protein levels of SAPCD2 in NPC cells. But SAPCD2 did not regulate the transcript level of PXN-AS1-L. Next, we investigated the detailed mechanism mediating the upregulation of SAPCD2 by PXN-AS1-L. PXN-AS1-L is mainly distributed in cytoplasm. Several cytoplasmic ln-cRNAs were shown to directly bind mRNAs and regulate the stability and/or translation of target mRNAs. 42,43 LncRNA BACE1-AS increased BACE1 mRNA stability and upregulated BACE1 protein expression. 42 Antisense Uchl1 was reported to promote UCHL1 mRNA translation. 43 In this study, we also revealed that PXN-AS1-L directly bound to SAPCD2 mRNA. Intriguingly, the interaction sites of SAPCD2 mRNA are located at 3′UTR. 3′UTR are well known target sites of miRNAs. Indeed, we found that the interaction between PXN-AS1-L and SAPCD2 mRNA decreased the binding of AGO2-miRNAs silencing complex on SAPCD2 mRNA. Dual luciferase reporter assays also showed that PXN-AS1-L increased SAPCD2 mRNA 3′UTR activity. Collectively, our findings suggested that PXN-AS1-L interacts with SAPCD2 mRNA 3′UTR and relieves the repressive roles of AGO2-miRNAs silencing complex on SAPCD2 mRNA stability and translation. The concrete miRNAs involved in the modulation need further investigation. Functional experiments further revealed that silencing of SAPCD2 significantly reversed the oncogenic roles of PXN-AS1-L in NPC in vitro and in vivo, which supported that SAPCD2 was an important mediator of the roles of PXN-AS1-L in NPC. In this study, we also found that PXN-AS1-L interacts with PXN mRNA as reported in HCC. 33 PXN may be another mediator of the roles of PXN-AS1-L in NPC, which needs further investigation. This study identified a novel action mechanism of PXN-AS1-L in NPC, which suggested the complex of action mechanisms of ln-cRNAs in different cancers. More completely investigating the molecular mechanisms of PXN-AS1-L will benefit the application of targeting PXN-AS1-L in cancer treatment. In this study, we focused on PXN-AS1-L. The expression, function, and action mechanism of another isoform PXN-AS1-S in NPC need further investigations to completely understand the significances of PXN-AS1.
In summary, this study found that lncRNA PXN-AS1-L is increased in NPC and correlated with poor prognosis of NPC patients. PXN-AS1-L promotes NPC cell proliferation, migration, and invasion in vitro, and NPC tumor growth in vivo via upregulating SAPCD2 expression. Targeted inhibition of PXN-AS1-L may be a potential anticancer strategy for NPC.