LncRNA EIF3J-AS1 enhanced esophageal cancer invasion via regulating AKT1 expression through sponging miR-373-3p

Esophageal cancer (ECa) remains a major cause of mortality across the globe. The expression of EIF3J-AS1 is altered in a plethora of tumors, but its role in ECa development and progression are undefined. Here, we show that EIF3J-AS1 is up-regulated in ECa and that its expression correlates with advanced TNM stage (P = 0.014), invasion depth (P = 0.001), positive lymph node metastasis (P < 0.001) and poor survival (OS: P = 0.0059; DFS: P = 0.0037) in ECa. Functional experiments showed that knockdown EIF3J-AS1 inhibited ECa growth and metastasis through in vitro and in vivo experiments. Regarding the mechanism, EIF3J-AS1/miR-373-3p/AKT1 established the ceRNA network involved in the modulation of cell progression of ECa cells. Overall, EIF3J-AS1 may exhibit an oncogenic function in ECa via acting as a sponge for miR-373-3p to up-regulate AKT1 mRNA level, and may serve as a potential therapeutic target and a prognostic biomarker for ECa patients.

Reporter assay for luciferase. Culturing of ECa cell was done in plates of 24 wells. After incubating for one full day, transfection of the reporter vector pmirGLO from Promega harboring the WT (wild-type) or MUT (mutant) EIF3J-AS1 was done into ECa cell combined with miR-373-3p. The miR-373-3p and AKT1 relationship was explored by culturing ECa cell in plates of 24 wells. After incubation for one full day, the transfection of the reporter vector pmirGLO harboring ZEB1 (WT or MUT) was done into ECa cell combined with miR-373-3p. Post-transfection of two full days, estimation of luciferase activity was done through Reporter System of Dual-luciferase from Promega.
In vivo metastasis assays. In vivo assessments were performed as previous 20 in male nude mice aged 6 weeks (Beijing Vitonlihua Experimental Animal Technology Co. Ltd, Beijing, China). Animals were housed in specified cages that were approved by the national animal guidelines of our institute. All animal experiments were approved by ethics committee of Zhejiang Cancer Hospital. 1 × 10 6 indicated Eca cells were subcutaneously injected into the flanks of 5 mice from (n = 5 for each group). The tumor volumes and weights were measured at the indicated time points (7, 14, 21, and 35 days). Five weeks following injection, mice were humanely killed by CO2 in accordance with ethical study requirements. For in vivo metastatic growth in the lungs, Mice were injected with indicated treated Eca cells (4 × 10 5 cells, 5 mice per group) in the tail vein to produce the pulmonary metastasis model. Ten weeks following injection, mice were humanely killed by CO2 in accordance with ethical study requirements and H&E stained to identify the presence of metastatic foci in the lungs. None anaesthetics were used during animal experiments.

Statistical analysis.
Data are shown and the mean ± SD analyzed via SPSS 17.0. Inter-group differences were assessed via a student's t-test. Multiple groups were compared using a one-way ANOVA analysis of variance. Kaplan Meier (KM) curves were plotted for survival analysis, and groups were compared via log-rank assessments (n = 3 for all). P-values < 0.05 were deemed significant differences.

EIF3J-AS1 is increased in ECa.
From the GEPIA (Gene Expression Profiling Interactive Analysis) data indicated that EIF3J-AS1 as increased in ECa tissues compared to normal tissues (Fig. 1A, Fold change > 1.5; p-value < 0.05). We next investigated EIF3J-AS1 levels in ECa vs. non-cancer cells and tissues via qRT-PCR analysis. EIF3J-AS1 expression was found to be higher in ECa tissues compared to healthy tissues ( Fig. 1B) with 64/182(35.16%) of ECa samples showing higher levels of the lncRNA (Fig. 1C). RT-PCR analysis further showed that EIF3J-AS1 was up-regulated in ECa cell lines vs. non-ECa HECC cells. Amongst the cell-types, EIF3J-AS1 showed the highest levels of expression in TE-1 and TE-8 cells (Fig. 1D). These data confirmed that EIF3J-AS1 expression is enhanced in ECa tissues and cells. www.nature.com/scientificreports/

High levels of EIF3J-AS1 expression correlate with aggressive clinicopathologic features and poor survival in ECa.
To investigate the correlation of EIF3J-AS1 expression and the clinicopathological characteristics of ECa, subjects were divided into high-and low-subgroups according to the median EIF3J-AS1 expression levels measured in ECa tissues (as a cut-off value). Table 1 shows that high levels of EIF3J-AS1 positively correlated with advanced TNM staging (P = 0.014), invasion depth (P = 0.001) and lymph node metastasis (P < 0.001). To explore the interplay between EIF3J-AS1 expression and the prognosis of ECa patient's, follow-up data was compiled. KM curves suggested that high levels of EIF3J-AS1 were associated with a poor ECa prognosis, poor OS, and low rates of DFS (Fig. 1E,F). These data confirmed that EIF3J-AS1 expression correlates with pathological stage, distant metastasis and survival in ECa patients.

Effects of EIF3J-AS1 Knockdown on ECa cells growth and metastasis.
To investigate the biology function role of EIF3J-AS1in ECa, we chose TE-1and TE-8 cells which have relative higher level of EIF3J-AS1 for further experiments. Through shRNA, sh-EIF3J-AS1#2 expression was significantly decreased in these cell lines ( Fig. 2A). CCK8 assay showed EIF3J-AS1 downregulation impaired the proliferation of TE-1and TE-8 cells (Fig. 2B). Next, cell migration and invasion were assessed by Transwell assay. Results showed that EIF3J-AS1 knockdown reduced migration and invasion of TE-1and TE-8 cells (Fig. 2C,D). Following, we injected indicated treated EcaTE-1 cells (sh-NC vs. sh-EIF3J2-AS1group) into the lateral tail vein of nude mice, the results indicating that knocking EIF3J2-AS1 group showed fewer numbers of metastatic foci and higher survival rate compared to sh-NC group (Fig. 2E-G). Meanwhile, tumor formation was delayed in sh-EIF3J2-AS1group knockdown group compared to sh-NC group (Fig. 2H,I). Taken together, above results indicate that EIF3J-AS1 was an oncogene in ECa.

EIF3J-AS1-mediated ECa metastasis via miR-373-3p/AKT1 axis. To study the regulation of ECa via
EIF3J-AS1, we found that EIF3J-AS1 was positively co-expression with AKT1 from the GEPIA (Gene Expression Profiling Interactive Analysis) data (Fig. 4A). Meanwhile, in ECa cells knock down EIF3J-AS1, a marked decrease in AKT1 mRNA level was observed (Fig. 4B). Furthermore, targetscan3.0 identified miR-373-3p as a putative binding target for AKT1 based on sequence complementarity (Fig. 4C), indicating that EIF3J-AS1 may act as ceRNA to sponge miR-373-3p to enhance to AKT1 inducing ECa progression. We then used a luciferase reporter assay approach to determine whether miR-373-3p was able to bind to its predicted target sequence in the AKT1 3′-untranslated region (UTR). We found that miR-373-3p mimic transfection was able to inhibit the activity of a wild-type (WT) AKT1 3′-UTR luciferase reporter, whereas it had no effect on a reporter in which its putative binding site had been mutated, consistent with the ability of miR-373-3p to directly bind to this 3′-UTR region (Fig. 4D). To explore whether EIF3J-AS1 regulates AKT1 expression through sponging miR-373-3p, www.nature.com/scientificreports/ sh-EIF3J plus miR-373-3p inhibitor or inhibitor NC was introduced into ECa cells. First, the efficiency of miR-373-3p inhibitor transfection was verified using RT-qPCR. The data revealed that transfection with miR-373-3p inhibitor resulted in a significant decrease in miR-373-3p expression in ECa cells (Fig. 4E). Furthermore, the downregulation of AKT1 mRNA (Fig. 4F) caused by EIF3J-AS1 knock down was reversed in ECa cells through miR-373-3p inhibitor re-introduction. To explore whether EIF3J-AS1 regulates ECa cells migration and invasion through AKT1, sh-NC, sh-EIF3J-AS1 or EIF3J-AS1 + LV-AKT1 was introduced into ECa cells. First, the efficiency of LV-AKT1 transfection was verified using RT-qPCR. The data revealed that transfection with LV-AKT1 resulted in a significant increase in AKT1 expression in ECa cells (Fig. 4G). Furthermore, the decreased migration and invasion ability of ECa cells caused by EIF3J-AS1 knock down was reversed in ECa cells through up-regulated AKT1or decreased AKT1 via miR-373-3p inhibitor (Fig. 4H). Taken together, EIF3J-AS1-Mediated ECa progression by via miR-373-3p/AKT1 axis.

Discussion
The dysregulation of lncRNAs is increasing linked to tumorigenesis, with both oncogenic and tumor suppressor functions reported [22][23][24][25] . In this study, higher levels of EIF3J-AS1 expression were observed in ECa tumors and cells, which was associated with advantaged TNM staging (P = 0.014), invasion depth (P = 0.001), metastasis (P < 0.001), and poor survival. Moreover, EIF3J-AS1 enhanced the proliferation and metastatic phenotypes of ECa cells through its ability to sponge miR-373-3p to increase AKT1 level. These findings collectively highlight EIF3J-AS1 is a key regulator of ECa tumorigenesis. LncRNAs act as quenchers of miRNA to regulate tumor progress. Currently, LncRNAs act as quenchers of miRNA to regulate tumor progress 26 . For example, the proliferative, invasiveness, and migratory capacity of pancreatic cancer cell lines are regulated by LncRNA linc00673 by sequestering miR-504 27 . Previous study indicated that EIF3J-AS1 may induce HCC progression via sponging miR-122e-5p 15 . Following, we wonder whether EIF3J-AS1 exerts roles through sponging other miRNAs to regulate AKT1 level. In our findings revealed a ceRNA model including EIF3J-AS1, miR-373-3p, and AKT1 in ECa cells. MiR-373-3p, transcribed from chromosome 19q13.42, with the function of cancer inhibition or promotion, belongs to the miRNAs-371-372-373 family and is commonly deregulated in many cancers 28 . And recently, it is proven that miR-373-3p plays a vital role in the regulation of breast cancer 29,30 , testicular germ cell tumors 31 , etc. Nevertheless, the role and inherent mechanisms of miR-373-3p in human Eca development remains to be assessed. Here, we could confirm that the one of targets of miR-373-3p is AKT1. We could prove the suppression of AKT1 by miR-373-3p in ECa cell lines. AKT1 was an oncogene in several cancer diseases including osteosarcoma, pancreatic ductal adenocarcinoma, colorectal cancer, and breast cancer [32][33][34][35] .Recently, lncRNAs have been implicated in the miRNA/AKT1 axis modulation in www.nature.com/scientificreports/ human cancers. For instance, the LINC00324 counteract miR-615-5p and enhance the expression of AKT1 in lung adenocarcinoma 36 . The lncRNA RNCR3 participates in the malignant transformation of colorectal cancer through miR-1301-3p/AKT1 axis interaction 37 . The current data revealed that AKT1 expression was enhanced by EIF3J-AS1 in ECa cell lines by sequestering endogenous miR-373-3p. The rescue experiments further showed the modulation of malignancy, growth, and metastatic related traits of ECa cells by EIF3J-AS1 through the miR-373-3p/AKT1 axis. Our study indicated a novelty EIF3J-AS1/miR-373-3p/AKT1 axis in ECa progression. To summarize, our study proves the overexpression of lncRNA EIF3J-AS1 in ECa of humans. The ECa cell lines advancement and tumorigenesis are promoted by EIF3J-AS1 by miR-373-3p sponging and positively regulating AKT1 expression. This study shows a fresh perspective on the molecular basis for comprehending ECa progression in terms of the oncogenic characteristic of EIF3J-AS1.  In contrast, no changes were observed upon EIF3J-AS1 mimic transfection in cells co-transfected with a reporter in which these 3′-UTR binding sites had been mutated. (E, F) miR-373-3p and AKT1 mRNA was performed by RT-qPCR in indicated ECa cells. (G) AKT1 mRNA was performed by RT-qPCR in indicated ECa cells. (H) Transwell assays were performed to assess the migratory and invasive abilities of Indicated ECa cells. *P < 0.05; **P < 0.01; ***P < 0.001.