Targeting lncRNA PSMA3-AS1, a Prognostic Marker, Suppresses Malignant Progression of Oral Squamous Cell Carcinoma

Objective Oral squamous cell carcinoma (OSCC) represents the most common maxillofacial malignancy. This study elucidated the clinicopathological value and molecular mechanisms of PSMA3 antisense RNA 1 (PSMA3-AS1) in OSCC. Methods Totally, 135 OSCC patients were recruited. PSMA3-AS1 expression and its prognostic value were assessed in this cohort. si-PSMA3-AS1 was transfected into HN4 and CAL-27 OSCC cells. Then, cell proliferation was evaluated by CCK-8, colony formation, and EdU staining. Migration and invasion were investigated through wound healing, transwell, and western blot. The PSMA3-AS1/miR-136-5p and miR-136-5p/FN1 interactions were validated by dual luciferase report, real-time quantitative polymerase chain reaction (RT-qPCR), and western blot. Results PSMA3-AS1 upregulation was determined in OSCC tissues. The upregulation indicated pessimistic patients' outcomes. Multivariate Cox regression analyses confirmed PSMA3-AS1 as an independent prognostic indicator. Its upregulation was also found in OSCC cells. Under transfection with si-PSMA3-AS1, proliferation, migration, and invasion were all restrained in HN4 and CAL-27 OSCC cells. Furthermore, its knockdown induced the increase in E-cadherin expression and the reduction in N-cadherin and Vimentin expression. PSMA3-AS1 was a sponge of miR-136-5p. Mutual inhibition was found between two and the interactions were confirmed by dual luciferase report. It was confirmed that FN1 was a target of miR-136-5p. FN1 expression was increased by miR-136-5p inhibitors, which was lessened by si-PSMA3-AS1 cotransfection. Conclusion Collectively, PSMA3-AS1 as a risk factor facilitated malignant behaviors of OSCC cells, related to the miR-136-5p/FN1 axis. Hence, PSMA3-AS1 as a potential therapeutic target for OSCC deserved further exploration.


Introduction
Oral squamous cell carcinoma (OSCC) represents the most common maxillofacial malignancy worldwide, accounting for approximately 90% among head and neck malignancies [1]. More than 300,000 new cases occur each year globally [2]. Meanwhile, death cases exceed 140,000 each year. The five-year survival rate is <50% [3]. The main therapeutic choices contain surgery, radiotherapy, and chemotherapy [4]. However, OSCC exhibits high invasiveness, low response to antitumor therapies, early recurrence, and undesirable clinical outcomes [5]. Putting together, it requires discovering innovative molecular targets and developing new therapeutic markers that possess higher potency.
lncRNA is a noncoding RNA with >200 nucleotides in length [6]. Studies have shown that lncRNA participates in mediating cancer-related miscellaneous physiological as well as pathological processes [7]. Data show that during tumorigenesis process, lncRNA as a competitive RNA (ceRNA) competitively binds to miRNA, thereby regulating and interacting with each other as well as affecting downstream mRNA expression [8]. As reported in previous research, PSMA3-AS1 functions as an oncogenic lncRNA [9]. PSMA3-AS1 upregulation may facilitate the progress of multiple myeloma [9], esophageal cancer [10], lung cancer [11], colorectal cancer [12], and glioma [13]. Mechanically, PSMA3-AS1 accelerates lung cancer growth and invasion through sponging miR-4504 [11]. PSMA3-AS1 as a sponge of miR-4429 induces colorectal cancer cell migration and invasion [12]. PSMA3-AS1 facilitates malignant phenotypes of esophageal cancer via the miR-101/EZH2 axis [10]. Nevertheless, it remains undiscovered concerning the roles and mechanisms of PSMA3-AS1 during OSCC. Here, we aimed to observe the roles of PSMA3-AS1 on OSCC progress and assess the potential of PSMA3-AS1 expression as an alternative marker for accurately predicting outcomes and as an underlying target for OSCC therapy. Moreover, this study observed the underlying mechanisms of PSMA3-AS1 in OSCC cells.

Materials and Methods
2.1. Patients and Specimens. Totally, 135 patients who underwent radical resection of OSCC in Affiliated Hospital of Weifang Medical University from January 2013 to August 2014 were selected as the research objects. The inclusion criteria were as follows: (1) pathological examination that confirmed the diagnosis; (2) first diagnosis; (3) no radiotherapy, chemotherapy, or other treatment; and (4) complete clinical and follow-up information. The normal oral mucosa tissue over 2 cm from the edge of the tumor was selected as the control. Fresh tissues that were frozen at −80°C after the operation were used for further analyses. The study was approved by the Ethics Committee of the Affiliated Hospital of Weifang Medical University (2013005), and an informed consent form was signed with the patient or family member.

Real-Time Quantitative Polymerase
Chain Reaction (RT-qPCR). The total RNA of tissue or cell specimens was extracted, and the extraction method was carried out according to the Trizol operating instructions (Beyotime, Shanghai, China). The reverse transcription reaction system contained 1 μL RT Enzyme Mix I, 1 μL Random 6 mers, 4 μL 5x Prime-Script Buffer, 1 μL Oligo dT Primer, and RNase-free H 2 O. cDNA was collected and PCR reaction was performed with SYBR. Table 1 lists the primers for PSMA3-AS1, miR-136-5p, GAPDH, and U6. The PCR reaction program included predenaturation at 95°C for 30 s, denaturation at 95°C for 5 s, and annealing extension at 60°C for 60 s, totally 40 cycles. The result was calculated on the basis of the 2 -ΔΔCt method.

Cell Counting
2.6. Colony Formation Assay. HN4 and CAL-27 cells were digested with 0.25% trypsin into a single cell suspension. 500 cells transfected with si-NC or si-PSMA3-AS1 were seeded in a 60 mm cell culture dish. After 3 days of continuous cultivation, the culture solution was changed. 3 mL culture medium was added to each dish and continued to culture for 8 days. Then, the culture was terminated. Cells were stained by Giemsa after the cells were fixed. Pictures were taken and the cell colonies were counted. 2.7. 5-Ethynyl-2 ′ -deoxyuridine (EdU) Staining. HN4 and CAL-27 cells transfected with si-NC and si-PSMA3-AS1 were seeded on 96-well plates (8 × 10 3 cells/well) and cultured overnight. EdU solution (Sigma, USA) was diluted with complete medium at a ratio of 1000 : 1 to prepare an appropriate amount of 50 μmol/L EdU medium. 100 μL 50 μmol/L EdU medium was added to each well for 2 h incubation. The culture medium was discarded, and the cells were washed with PBS buffer twice for 5 min each time. 100 μL cell fixative was added to each well and incubated for 30 min at 37°C. Following adding 2 mg/mL glycine, cells were incubated for 5 min on a decolorizing shaker. 100 μL PBS buffer was added to each well and continued to incubate for 5 min on a decolorizing shaker. After adding 100 μL 1x Apollo staining reaction solution to each well, cells were incubated for 30 min at 37°C in the dark. The staining reaction solution was removed, and 100 μL penetrant was added. Samples were washed twice on a decolorizing shaker, each 2.9. Transwell Assay. Transwell assays were used for measuring migration as well as invasion. Matrigel (Corning, USA) was used for coating before the invasion experiment, and the coating step was omitted before the migration experiment. The rest of the steps was the same and was briefly described as follows: the cells in si-NC and si-PSMA3-AS1 groups were suspended in a serum-free cell culture medium. Then, 200 μL cells were drawn and planted in the upper chamber. At the same time, 500 μL serum-containing cell culture medium was added to the lower chamber. The chamber was placed in a CO 2 incubator at 37°C for 24 h. The cells were fixed with methanol and stained with crystal violet. Five fields were randomly selected under the microscope (Olympus, Japan) to count the number of cells penetrating the membrane.
2.10. Dual Luciferase Report. The wild-type (WT) or mutant (MUT) gene sequences were cloned into the pmirGLO plasmid to obtain pmirGLO-PSMA3-AS1 WT/MUT and pmirGLO-FN1 WT/MUT. Using Lipofectamine™ 2000 transfection reagent, the recombinant plasmids combining with miR-NC or miR-136-5p mimics were transferred into HN4 or CAL-27 cells. After 48 h, the luciferase activity was detected according to the steps described in the dual luciferase reporter kit (Beyotime, Shanghai, China).
2.11. Western Blot. Western blot method was used for determining E-cadherin, N-cadherin, Vimentin, and FN1 proteins. The steps were briefly described as follows: the total protein was extracted from cells in each group. The amount of protein loaded in each well was 40 μg. The electrophoresis voltage was 90 V, and the electrophoresis at constant voltage was 0.5 h. Under electrophoresis at 120 V, constant voltage lasted for 2 h. The NC membrane was cut to the same size as the separation glue. The membrane was transferred under the condition of 60 V constant pressure; the transfer device was performed at 4°C. And the transfer time was set to 60 min. The membrane was closed by 5% skimmed milk powder. The membrane was incubated with primary antibodies against E-cadherin (1 : 1000; #3195; CST, USA), N-cadherin (1 : 1000; #13116; CST), Vimentin (1 : 1000; #5741; CST), FN1 (1 : 1000; #26836; CST), and GAPDH (1 : 1000; #5174; CST) overnight and incubated with secondary antibodies (#7075; CST) for 2 h. Color was developed with electrochemical luminescence method (Millipore, USA). As a result, ImageJ was used for analyzing the OD value of the band with GAPDH as an internal reference.
2.12. Statistical Analysis. GraphPad Prism 5, SPSS 23.0, and R software were utilized for statistical analysis. Quantitative data were expressed by mean ± standard deviation. Comparisons between groups were assessed via Student's t-test or one-way analysis of variance. Correlation between PSMA3-AS1 expression and clinicopathological parameters was analyzed through the chi-square test. Kaplan-Meier overall survival (OS) and disease-free survival (DFS) curves were depicted between high and low PSMA3-AS1 expression subgroups. The differences in survival between subgroups were confirmed through log-rank test. Multivariate Cox regression analyses were used for assessing which parameters independently predicted subjects' survival. P value < 0.05 was considered statistically significant.

Results
3.1. lncRNA PSMA3-AS1 Upregulation as a Prognostic Factor of OSCC. Here, 135 OSCC subjects were recruited. PSMA3-AS1 overexpression was displayed in OSCC than normal tissue specimens (Figure 1(a)), indicating that this lncRNA might be in relation to OSCC progress. We compared PSMA3-AS1 expression in NHOK normal oral mucosal epithelial cells with four OSCC cell lines HN4, HN6, SCC-25, and CAL-27 ( Figure 1(b)). Data were indicative of the upregulation of PSMA3-AS1 in OSCC than normal cell lines. Then, the prognostic value of PSMA3-AS1 upon OSCC was under evaluation. Table 2 lists that PSMA3-AS1 expression did not exhibit distinct associations with gender, age, and tumor size among OSCC subjects. Nevertheless, PSMA3-AS1 expression was in relation to distant metastasis as well as TNM stage, which was indicative that this lncRNA might participate in malignant progress of OSCC. We assessed whether PSMA3-AS1 expression was predictive of subjects' outcomes. In Figure 1(c), PSMA3-AS1 upregulation was indicative of shorter OS time in comparison to its downregulation. In the meanwhile, subjects with its low expression experienced pessimistic DFS time (Figure 1(d)).
To verify whether this lncRNA was independently predictive of OSCC outcomes, multivariate Cox regression analyses were carried out. Following incorporating gender, age, tumor size, distant metastasis, TNM stage, and PSMA3-AS1 expression, our data confirmed that TNM stage and PSMA3-AS1 expression were both independently predictive of subjects' outcomes (Table 3). Collectively, PSMA3-AS1 upregulation was a prognostic indicator of OSCC.

PSMA3-AS1 Knockdown Weakens OSCC Cellular
Proliferative Potential. This study explored the molecular basis concerning high PSMA3-AS1 expression for promotion of OSCC subjects' poor outcomes. To silence PSMA3-3 Disease Markers AS1, siRNAs against PSMA3-AS1 were transfected into HN4 and CAL-27 OSCC cells. The transfection results were examined via RT-qPCR. In Figure 2(a), PSMA3-AS1 displayed lowered expression in si-PSMA3-AS1-transfected HN4 and CAL-27 OSCC cells, indicating that this lncRNA was successfully silenced. Whether PSMA3-AS1 affected OSCC cellular proliferation was under assessment. CCK-8 assay was utilized for cell viability at 0, 24, 48, 72, and 96 h after transfection with si-PSMA3-AS1 or si-NC. In Figure 2(b), silencing PSMA3-AS1 exhibited the restrained functions on HN4 cellular viability. The similar data were observed in CAL-27 OSCC cells. Clone formation capacities of OSCC cells were then assessed. In comparison to si-NC transfection, si-PSMA3-AS1 transfection displayed the decreased colony number of HN4 cells as well as CAL-27 cells (Figure 2(c)). EdU staining was used for assessing cellular proliferation. As depicted in Figure 2(d), proliferative abilities were markedly lessened by PSMA3-AS1 knockdown. Hence, PSMA3-AS1 exerted a stimulative role on OSCC proliferation.

3.3.
Silencing PSMA3-AS1 Lessens OSCC Cellular Migratory and Invasive Abilities. Migration and invasion are key manifestations of the malignant behaviors of OSCC. This study evaluated whether PSMA3-AS1 influenced migratory as well as invasive abilities in OSCC cells. According to wound healing assay data, the scratch of si-PSMA3-AS1-transfected HN4 and CAL-27 cells showed wider distance than si-NC transfection (Figure 3(a)). The data were indicative of the   Figure 3(b), compared to the si-NC group, the si-PSMA3-AS1 group exhibited the lowered number of invasive HN4 and CAL-27 cells. Also, the migratory OSCC cell number had a distinct reduction by PSMA3-AS1 knockdown (Figure 3(c)). Epithelial-tomesenchymal transition (EMT) process exerts a key function in cancer migration as well as metastases. Here, E-cadherin, N-cadherin, and Vimentin proteins were detected in HN4 and CAL-27 cells through western blot. Consequently, si-PSMA3-AS1 transfection markedly augmented E-cadherin expression as well as lessened N-cadherin and Vimentin expression in OSCC cells (Figure 3(d)). This indicated that EMT process of OSCC cells was obstructed by silencing PSMA3-AS1. Taken together, PSMA3-AS1 knockdown weakened migratory as well as invasive capacities of OSCC cells.

Discussion
In this study, we proposed a novel lncRNA PSMA3-AS1 that exhibited upregulation in OSCC. It acted as an independent risk factor for this malignancy. Moreover, it was involved in OSCC malignant biological properties containing prolifera-tion, migration, and invasion, which was in relation with the miR-136-5p/FN1 axis.
OSCC subjects' treatment is mainly on the basis of clinicopathological indicators like TNM and grade [15]. Nevertheless, it has been under confirmation that novel molecular biomarkers may prolong survival time. Thus, it requires probing novel molecules that may characterize subjects' risk stratifications. Here, PSMA3-AS1 showed upregulation in OSCC than normal specimens. Its expression could be utilized for independently predicting patients' outcomes. Previously, Xu et al. reported that circulating PSMA3-AS1    Disease Markers exhibited distinct correlations to multiple myeloma subjects' prognosis [9]. Qiu et al. found that elevated PSMA3-AS1 expression was in relation with tumor size, distant metastases, and pessimistic outcomes for esophageal cancer [10]. Li et al. demonstrated that its expression showed positive correlations to stage, metastasis, and poor outcomes for lung cancer [11]. As reported in the study of Wang et al., PSMA3-AS1 expression was in correlation to stage, lymph node metastases, and OS in non-small-cell lung carcinoma [16]. PSMA3-AS1 was a prognostic marker for diffuse large Bcell lymphoma [17]. Thus, PSMA3-AS1 may be an oncogene for OSCC. Our data confirmed that targeting PSMA3-AS1 weakened OSCC cell proliferative characteristics. Lymph node metastasis is a critical clinical feature of OSCC, as the primary cause of postoperative recurrence in OSCC patients [18]. Tumor cell metastasis contains the corresponding sequence and related steps. Many of these steps are affected by EMT. The ability of cells to obtain mesenchymal properties enhances cell migration and invasion, induces the stem cell characteristics of tumor cells, and inhibits cell apoptosis and programmed senescence, allowing cells to metastasize [19][20][21]. EMT is a key for therapy resistance, recurrence, and distant metastases [22]. The main characteristics of EMT in tissues or organs include the decrease in Ecadherin expression and the increase in Vimentin and Ncadherin expression [23]. Here, targeting PSMA3-AS1 restrained migratory and invasive features of OSCC cells, which elevated E-cadherin level as well as lessened Vimentin and N-cadherin levels in OSCC, thereby inactivating EMT process.
Several limitations should be pointed out. First, the prognostic value of PSMA3-AS1 expression should be verified in a prospective cohort. Second, it requires investigating the carcinogenic role of PSMA3-AS1 in vivo. Third, more experiments will be presented for confirming the interactions of PSMA3-AS1/miR-136-5p/FN1.

Conclusion
Collectively, in the OSCC cohort, PSMA3-AS1 was independently predictive of patients' outcomes. PSMA3-AS1 upregulation was implicated in carcinogenesis as well as cancer cellular proliferative, migratory, and invasive features. In terms of mechanisms, PSMA3-AS1 could play a carcinogenic effect through the miR-136-5p/FN1 axis. Hence, targeting carcinogenic PSMA3-AS1 offered novel therapeutic thought against OSCC.

Data Availability
The datasets analyzed during the current study are available from the corresponding author on reasonable request.