lncRNA TPT1-AS1 promotes cell migration and invasion in esophageal squamous-cell carcinomas by regulating the miR-26a/HMGA1 axis

Abstract lncRNA TPT1-AS1 plays an oncogenic role in ovarian and cervical cancers. However, its involvement in the pathological progress of esophageal squamous-cell carcinomas (ESCCs) is unclear. lncRNA TPT1-AS1 was mainly localized in the cytoplasm of ESCC cells and interacted with miR-26a. In ESCC tissues, lncRNA TPT1-AS1 level was obviously increased, while miR-26a level was decreased. Interestingly, lncRNA TPT1-AS1 level was not significantly correlated with miR-26a level but was positively correlated with HMGA1 mRNA, a target of miR-26a. In ESCC cell lines KYSE510 and KYSE-30, lncRNA TPT1-AS1 overexpression enhanced HMGA1 expression, while it had no effect on miR-26a expression. Cell migration and proliferation assays indicated that lncRNA TPT1-AS1 and HMGA1 overexpression promoted ESCC cell migration and invasion, while their effects were alleviated by miR-26a overexpression. The migration and invasion of ESCC cells were suppressed by lncRNA TPT1-AS1 knockdown. In conclusion, lncRNA TPT1-AS1 plays an oncogenic role in ESCC and might function by upregulating HMGA1 via sponging miR-26a.


Introduction
Esophageal cancer has been estimated to affect 572,034 new cases and has caused 508,585 deaths worldwide in 2018 [1]. Esophageal squamous-cell carcinoma (ESCC) is the common type of esophageal cancer [2]. Smoking, heavy alcohol consumption, and human papillomavirus infection are the main risk factors of ESCC [3,4]. Despite the continuous improvement in the medical standards in recent years, the 5 year survival rate of patients with esophageal cancer is not ideal [5][6][7]. Therefore, it is still necessary to develop novel diagnostic markers and therapeutic strategies.
It has been well established that the development of ESCC involves multiple molecular pathways [8]. The functional molecules involved in these pathways may be potential diagnostic markers and therapeutic targets [9,10]. Long (>200 nt) non-coding RNAs (lncRNAs) can function as gene regulators by interacting with DNA (e.g., promoters), RNA, or proteins [11]. Recently, mounting lncRNAs have been recognized as important regulators in ESCC. Accumulating evidence has proposed that lncRNAs play significant roles in the biological development of ESCC by regulating cell proliferation and apoptosis [12,13]. For instance, lncRNA small nucleolar host gene 1 (SNHG1), SNHG6, and SNHG16 have been reported to promote ESCC cell proliferation [14,15]. Cox univariate and multivariate analyses have revealed that SNHG1 is an independent prognostic factor for overall survival (OS) and disease-free survival in ESCC patients. The interaction between miRNA-21 and SNHG1 plays an important role in ESCC cell proliferation [16]. lncRNA SNHG1 promotes the development of cervical cancer cells. lncRNA SNHG1 knockdown decreases the proliferation, migration, and invasion of HeLa and C-33A cells [17]. lncRNA SNHG1 is upregulated in ESCC tissues and high SNHG1 expression is positively correlated with ESCC lymph node metastasis and decreased OS. CASC9 promotes ESCC growth by regulating the EZH2/CASC9 pathway [18], while KLF3-AS1 inhibits ESCC tumorigenesis by regulating the miR-185-5p/KLF3 axis [19]. TPT1-AS1, a lncRNA located on chromosome 13 with about 76,000 base pairs, has an oncogenic role in ovarian and cervical cancers [20][21][22]. Moreover, TPT1-AS1 accelerates the progression of colorectal cancer (CRC) by upregulating TPT1 levels and activating the FAK and JAK-STAT3 signaling pathways [23]. Tiang et al. has studied TPT1-AS1 and reported that TPT1-AS1 silencing suppressed gastric cancer development [22]. However, the effect of TPT1-AS1 on ESCC remains largely unknown. TPT1-AS1 is predicted to interact with miR-26a by IntaRNA software. Existing studies have revealed that miR-26a and its target HMGA1 are involved in the pathological progress of many cancers [24][25][26]. For instance, the interaction between miR-26a and its target gene HMGA1 might contribute to the malignant progression of human urothelial bladder cancer [27]. In addition, miR-26a-5p overexpression could be a novel therapy to improve coronary microembolization-induced myocardial damage. Studies have shown that HMGA1 is a target gene of miR-26a-5p. However, whether the TPT1-AS1/miR-26a axis plays a role in ESCC remains unclear. Here we attempted to verify the relationship between TPT1-AS1 and miR-26a and investigate the potential roles of TPT1-AS1 and miR-26a/HMGA1 axis in ESCC.

Research subjects
60 ESCC patients from Shanxi Cancer Hospital (Shanxi, China) between January 2017 and January 2019 were involved. Patients who had a history of malignancies or ESCC treatment or had other clinical diseases were excluded. The correlation between TPT1-AS1/miR-26a/ HMGA1 expressions and clinicopathological factors of ESCC patients are listed in Tables 1-3.
Ethics approval and consent to participate: The present study was approved by the Ethics Committee of Shanxi Cancer Hospital. The research has been carried out in accordance with the World Medical Association Declaration of Helsinki. All patients and healthy volunteers provided written informed consent prior to their inclusion within the study. The screening criteria were P < 0.05 and a fold change >1.
The study included 60 patients (29 males and 31 females). Among them, 30 patients were assigned in the high TPT1-AS1 expression group. Of these 30 patients, 15 patients had lymph node metastasis and 15 had no lymph node metastasis.

ESCC tissue samples and cells
Fine needle biopsies under the guidance of MRI were performed to collect ESCC tissues and their matched adjacent normal-appearing tissues. The tissue sections were stained by hematoxylin and eosin, and their histopathological features were examined by three professional pathologists. The ESCC cancer tissues were diagnosed according to the World Health Organization's classification for esophageal cancer [28]. The matched adjacent normal-appearing tissues were diagnosed with the criteria that there were no detectable cancer cells in the samples. KYSE510 and KYSE-30 ESCC cell lines were from BFB Biotechnology Co., Ltd (China). They were cultured in RPMI-1640 medium with 10% FBS under 5% CO 2 at 37°C. The cells were harvested from passages 3-6 for subsequent experiments. All experiments were performed with mycoplasma-free cells, which were authenticated by STR analysis.

Subcellular fractionation
The nuclear and cytoplasm extracts of KYSE510 cells were isolated using NE-PER Nuclear and Cytoplasmic Extraction Reagents kit (pierce-78835, Pierce, USA). The TPT1-AS1, U6, and GAPDH levels in the nuclear and cytoplasmic extracts were detected by RT-qPCR.

Western blotting
Total proteins were isolated from in vitro cultivated cells and quantified using bicinchoninic acid assay method. 50 μg (10 μL) of total proteins were loaded to each sample well and separated by 10% SDS-PAGE. The separated proteins were transferred onto PVDF membranes, which were blocked with 5% nonfat dry milk for 2 h and incubated with antibodies against GAPDH (ab9845, Abcam) and HMGA1 (ab226850, Abcam) at 4°C for 18 h, followed by incubation with secondary antibody (ab6721, Abcam) at room temperature for 2 h. The target protein signals were detected using the enhanced chemiluminescence system (Amersham).

Transwell assays
Transwell assays were performed using KYSE510 and KYSE-30 cells collected at 48 h post-transfection using Transwell inserts (8 µm, 3415, Corning). For invasion assay, the Transwell inserts were pre-treated with 50 μL/well Matrigel (dilution rate = 1:3) at 37°C for 6 h. For migration assay, the uncoated Transwell inserts were used. 4,000 cells were seeded into the upper chamber, and RPMI-1640 media with 20% FBS were added into the lower chamber. 12 h later, the lower surface of membranes was stained using crystal violet (0.1%, Sigma-Aldrich) for 20 min. Images were analyzed using ImageJ software to measure the percentage of the area of cells present in each insert.

Statistics
The experimental data (three replicates) were analyzed using SPSS 17.0 software. Data from ESCC tissue samples were analyzed using paired t test and Pearson's correlation coefficient. One-tailed t-test was used in the study. Unpaired t test was applied to compare the difference between two groups. ANOVA was used to compare the difference among multiple groups. The threshold of pvalue was set to 0.05.

TPT1-AS1 interacted with miR-26a
TPT1-AS1 was predicted to interact with miR-26a by IntaRNA ( Figure 1a). As shown in Figure 1b, miR-26a overexpression markedly inhibited the luciferase activity in KYSE510 cells transfected with pGL3-TPT1-AS1-promoter and pRL-TK, indicating that TPT1-AS1 interacted with miR-26a. Moreover, we isolated the nuclear and cytoplasmic extracts from ESCC cell line KYSE510. Subsequent RT-qPCR assay indicated that TPT1-AS1 was mainly localized in the cytoplasm of KYSE510 cells (Figure 1c). MiR-26a overexpression significantly reduced the luciferase activity of the HMGA1-wt vector but failed to decrease that of the HMGA1-mut ( Figure A1).
3.2 TPT1-AS1 level was increased while miR-26a level was reduced in ESCC tissues TPT1-AS1 and miR-26a levels in ESCC and matched nontumor tissues were assessed. The results indicated that TPT1-AS1 level was markedly increased in ESCC tissues (Figure 2a), while miR-26a level was apparently reduced in ESCC tissues (Figure 2b). The expression levels of TPT1-AS1 and miR-26a/HMGA1 in these two cells were compared with KYSE180 cells and normal cells. The results indicated that TPT1-AS1 level was increased in KYSE180 cells and KYSE510 cells ( Figure A3a), while miR-26a level was reduced in KYSE180 cells and KYSE510 cells ( Figure A3b). In addition, HMGA level was increased in KYSE180 cells and KYSE510 cells ( Figure A3c). Moreover, TPT1-AS1 level was increased in KYSE180 cells and KYSE-30 cells (Figure A3d), while miR-26a level was reduced in KYSE180 cells and KYSE-30 cells ( Figure A3e). Moreover, HMGA level was increased in KYSE180 cells and KYSE-30 cells ( Figure A3f).

TPT1-AS1 enhanced ESCC cell migration and invasion by regulating the miR-26a/ HMGA1 axis
The function of TPT1-AS1 in ESCC cells was explored. TPT1-AS1 and HMGA1 overexpression markedly promoted migration and invasion of ESCC cells, while miR-26a overexpression and TPT1-AS1 knockdown obviously suppressed migration and invasion of ESCC cells. Moreover, the effects of TPT1-AS1 and HMGA1 overexpression on migration and invasion were alleviated by miR-26a overexpression (Figure 5a and b).

Discussion
Our study mainly explored the role of TPT1-AS1 in ESCC. We revealed that TPT1-AS1 is remarkably increased in ESCC, and it could enhance ESCC cell migration and invasion by regulating the miR-26a/HMGA1 axis. Therefore, TPT1-AS1 might be a possible therapeutic target for ESCC. Previous studies have shown that TPT1-AS1 knockdown significantly inhibits proliferation and cell cycle G1/S transition of SGC-7901 and MGC-803 cells. TPT1-AS1 diminishes cell proliferation and sensitizes cells to chemotherapy by sponging miR-3156-5p and upregulating CASP2. In addition, it has been reported that TPT1-AS1 induces epithelial ovarian cancer (EOC) tumor growth and metastasis through TPT1 and downstream PI3K/AKT signaling and that TPT1-AS1 might be a promising therapeutic target for EOC. Recent studies have explored the involvement of TPT1-AS1 in ovarian and cervical cancers [20,21]. In ovarian cancer, TPT1-AS1 is upregulated and induces TPT1 expression to promote cancer cell metastasis [21]. In cervical cancer, TPT1-AS1 is increased and sponges miR-324-5p [30] to promote tumor metastasis and growth [20]. Previous studies have also shown that TPT1-AS1 enhances EOC cell proliferation, migration, and invasion via the TPT1/PI3K/AKT signaling pathway in vitro. In addition, TPT1-AS1 promotes tumor progression and metastasis in CRC by upregulating TPT1 level and activating the FAK and JAK-STAT3 signaling pathways. The phenotypic changes in pathways related to migration and invasion remain to be further elucidated [22,31]. Here we observed that TPT1-AS1 is remarkably increased in ESCC tissues, and its overexpression enhances ESCC cell migration and invasion. Therefore, TPT1-AS1 also has an oncogenic function in ESCC. Moreover, we uncovered that TPT1-AS1 interacts with miR-26a. Existing studies have found that the role of miR-26a in cancers is differentiated. In CRC, miR-26a is apparently downregulated and can regulate FUT4 expression to suppress cancer cell aggressiveness [32]. In triple-negative breast cancer, miR-26a is downregulated and can target metadherin to suppress cancer cell migration and proliferation [33]. However, miR-26a is increased in ovarian cancer and promotes cancer cell proliferation [34].
Here we observed that miR-26a is decreased in ESCC tissues, and its overexpression represses ESCC cell migration and invasion, consistent with a recent report [35].
Studies have found that miR-26a suppresses bladder cancer by regulating its target HMGA1 [26]. HMGA1 can regulate various genes due to its ability to alter chromatin structures. HMGA1 overexpression is a hallmark of human cancers and exhibits a pivotal role in cell transformation [36]. miR-26a downregulates HMGA1 by targeting its 3′-UTR, and HMGA1 knockdown significantly suppresses the migration and invasion of two osteosarcoma cell lines in vitro [37]. Moreover, we uncovered that TPT1-AS1 might play its oncogenic function by modulating the miR-26a/HMGA1 axis. Further studies are needed to explore other potential mechanisms.
Our study also has some limitations. First, our sample size of ESCC patients is small, and all patients are Han Chinese. Therefore, a bigger sample size with different ethnic backgrounds is needed to further confirm our conclusion. In addition, in vivo animal experiments are needed to assess the role of TPT1-AS1 in tumor metastasis. The effect of HMGA1 knockdown on ESCC cells migration and invasion should be confirmed in the future.

Conclusion
TPT1-AS1 level is markedly increased in ESCC tissues, and its overexpression enhances ESCC cell migration and invasion via modulating the miR-26a/HMGA1 axis. Conflict of interest: The authors declare no conflicting interests.

List of abbreviations
Data availability statement: The experimental data are available from the corresponding author on reasonable request. Figure A1: Luciferase assays in ESCC cells co-transfected with wildtype or mutant HMGA1 and miR-26a or NC. MiR-26a overexpression significantly reduced the luciferase activity of the HMGA1-wt vector.