The molecular mechanisms of the long noncoding RNA SBF2-AS1 in regulating the proliferation of oesophageal squamous cell carcinoma

The long noncoding RNASBF2-AS1 can promote the occurrence and development of many kinds of tumours, but its role in oesophageal squamous cell carcinoma (ESCC) is unknown. We found that SBF2-AS1 was up-regulated in ESCC, and its expression was positively correlated with tumor size (P = 0.0001), but was not related to gender, age, TNM stage, histological grade, and lymphnode metastasis (P > 0.05). It was further found that the higher the expression of SBF2-AS1, the lower the survival rate. COX multivariate analysis showed that the expression of SBF2-AS1 was an independent prognostic factor. Functional experiments show that inhibition of SBF2-AS1 can inhibit the proliferation of ESCC through in vivo and in vitro, and overexpression of SBF2-AS1 can promote the proliferation of ESCC and inhibit its apoptosis. In mechanism, SBF2-AS1/miR-338-3P, miR-362-3P/E2F1 axis are involved in the regulation of ESCC growth. In general, SBF2-AS1 may be used as ceRNA to combine with miR-338-3P and miR-362-3P to up-regulate the expression ofE2F1, and ultimately play a role in promoting cancer. It may be used as a therapeutic target and a biomarker for prognosis.


Scientific Reports
| (2021) 11:805 | https://doi.org/10.1038/s41598-020-80817-w www.nature.com/scientificreports/ in SUV12 and EZH2 was significantly decreased, resulting in the upregulation of P21 expression. Overexpression of SBF2-AS1 reduced the expression of P21 and promoted tumorigenesis 13 . CeRNAs are responsible for a new kind of regulatory mechanism between noncoding RNAs and messenger RNAs 14 . An increasing number of studies have found that many lncRNAs mainly act as ceRNAs to regulate the occurrence and development of tumours, such as osteosarcoma 15 , non-small-cell lung cancer 16 , colorecta lcancer 17 and glioma 18 . However, our mechanistic understanding of the network of ceRNAs that target SBF2-AS1 in oesophageal squamous cell carcinoma is not clear. In this study, SBF2-AS1 was used as a sponge to adsorb miR-338-3P and miR-362-3P, upregulating the expression of E2F1 and ultimately promoting the proliferation of oesophageal squamous cell carcinoma. Additionally, SBF2-AS1 could promote the proliferation of oesophageal squamous cell carcinoma in vivo and in vitro.

Results
The SBF2-AS1 is highly expressed in ESCC. SBF2-AS1 expression was examined in ESCC. The results showed that the expression level of SBF2-AS1 in ESCC tissues was higher than that in adjacent normal tissues (Fig. 1a). Additionally, SBF2-AS1 was found to be up regulated in ESCC cell lines (Fig. 1b). These results suggest that SBF2-AS1 is highly expressed in ESCC tissues and cells.
High levels of SBF2-AS1 expression correlate with clinicopathologic features and poor survival in ESCC. In order to study the relationship between the expression level of SBF2-AS1 and the clinical pathological characteristics of patients, patients were divided into SBF2-AS1 high expression group and SBF2-AS1 low expression group. The results showed that the expression level of SBF2-AS1 was positively correlated with tumor size (P = 0.0001). The higher the expression level of SBF2-AS1, the larger the tumor, but there were no statisti- www.nature.com/scientificreports/ cal differences in gender, age, TNM stage, histological grade, and lymph node metastasis (Table 1). And further research found that the expression level of SBF2-AS1 is negatively correlated with the overall survival time of the patient. The higher the expression level of SBF2-AS1, the shorter the overall survival time of the patient (Fig. 1c). COX multivariate analysis further shows that the expression level of SBF2-AS1 is an independent factor predicting the poor prognosis of patients with ESCC ( Table 2).
Overexpression of SBF2-AS1 promoted ESCC cell proliferation. To analyse the effect of SBF2-AS1 on oesophageal squamous cell carcinoma cell lines, ECA109 and TE-13 cells were transfected with OE-SBF2-AS1, PCDNA, si-SBF2-AS1 and si-NC. RT-PCR showed that the expression of SBF2-AS1 in the overexpression group was significantly higher than that in the control group, while that in the si-SBF2-AS1 group was significantly lower than that in the control group (Fig. 2a). The RTCA results showed that SBF2-AS1 significantly promoted the proliferation of oesophageal squamous cell carcinoma, while si-SBF2-AS1 inhibited the proliferation of oesophageal squamous cell carcinoma (Fig. 2b). EdU and colony formation assays further showed that SBF2-AS1 could promote the proliferation of oesophageal squamous cell carcinoma (Fig. 2c,d). Flow cytometry analysis showed that SBF2-AS1 could promote the transition from G1 phase to S phase, and flow cytometry to analyse cell apoptosis showed that SBF2-AS1 could inhibit the apoptosis of oesophageal cancer cells (Fig. 2e,f).

SBF2-AS1
can adsorb miR-362-3P and miR-338-3P as a ceRNA. To determine the molecular mechanism by which SBF2-AS1 regulates ESCC, we first confirmed that SBF2-AS1 is mainly located in the cytoplasm through FISH experiments (Fig. 3a). Then, the downstream target of SBF2-AS1 was predicted by using the starBase website (http://starb ase.sysu.edu.cn/), and SBF2-AS1 was found to bind miR-338-3P and miR-362-3P.  www.nature.com/scientificreports/ And we also detected the expression levels of miR-338-3P and miR-362-3P in ESCC lines, and found that miR-338-3P andmiR-362-3P were low expressed in ESCC lines (Fig. 3b). Many studies have reported that SBF2-AS1 can be used as a ceRNA to promote the occurrence and development of tumours. And whether it can bind to AGO2 protein is regarded as an important marker to play the role of ceRNA. To confirm that SBF2-AS1 can be used as a ceRNA to adsorb miR-338-3P and miR-362-3P, we used RNA immunoprecipitation (RIP) to detect whether SBF2-AS1 could bind the AGO2 protein in oesophageal cancer cells. The results showed that the AGO2 protein bound more SBF2-AS1 than IgG, suggesting that SBF2-AS1 can act as a ceRNA (Fig. 3c). Furthermore, we carried out RNA-pull down experiments with biotin-labelled miR-338-3P and miR-362-3P in ECA109 cells and found that biotin-labelled miR-338-3P and miR-362-3P could bind more SBF2-AS1 than that bound in the control group (Fig. 3d). Therefore, it was confirmed that SBF2-AS1 can be used as a ceRNA to adsorb miR-338-3P and miR-362-3P.

SBF2-AS1 can upregulate the expression of E2F1 by adsorbing miR-338-3P and miR-362-3P. To
show that SBF2-AS1 upregulates the expression of E2F1 by adsorbing miR-338-3P and miR-362-3P, we designed a series of cross-rescue experiments. Expression of the E2F1 protein in two oesophageal cancer cell lines, ECA109 andTE-13 cells, was detected by Western blot assay. The results showed that the overexpression of SBF2-AS1 promoted the expression of E2F1, while si-SBF2-AS1 inhibited the expression of E2F1. At the same time, the expression levels of Cyclind1 andP21, which are closely related to the cell cycle, was detected. The results showed that SBF2-AS1promoted the expression of Cyclind1 and inhibited the expression of P21 (Fig. 5a). Western blot analysis also proved that miR-338-3P and miR-362-3P could abrogate the upregulated expression of E2F1 and Cyclind1 and downregulated P21 expression caused by SBF2-AS1 (Fig. 5b). Colony-formation, EdU and RTCA rescue assays showed that miR-338-3P and miR-362 could attenuate the effect of SBF2-AS1 on cell proliferation ( Fig. 5c,d,f). QRT-PCR rescue experiment showed that miR-338-3P and miR-362-3P decreased the expression level of SBF2-AS1 (Fig. 5e). These findings suggest that SBF2-AS1, acting as a ceRNA, adsorbs miR-338-3P and miR-362-3P and upregulates E2F1 to promote the proliferation of oesophageal cancer cells.

SBF2-AS1ASO could inhibit the proliferation of oesophageal cancer cells in vivo.
To investigate whether SBF2-AS1 interference could inhibit the proliferation of ESCC in vivo, we designed SBF2-AS1 ASO, a siRNA with a therapeutic effect. We used male SPF-grade Balb/C nude mice to carry out a subcutaneous tumour-bearing experiment. The mice were divided into the ASONC group and ASO group. The results showed www.nature.com/scientificreports/ that the tumour volume in the ASO group was significantly smaller than that in the ASONC group (Fig. 6a). Staining for E2F1 by immunohistochemistry was further analysed in these two groups, and the staining intensity of E2F1 in the ASO group was found to be significantly lower than that in the ASONC group (Fig. 6b). Therefore, SBF2-AS1 ASO could inhibit the growth of oesophageal cancer in vivo.

Discussion
Eukaryotic cells can transcribe many types of RNA, including protein-encoding RNAs, short non-coding RNAs and long non-coding RNAs 19 . According to their location in the genome, lncRNAs are divided into five types: sense lncRNAs, antisense lncRNAs, bidirectional lncRNAs, intra-gene lncRNAs and genomic lncRNAs 20 . These RNAs are involved in many disease processes 21 and participate in the occurrence and development of many kinds of malignant tumours and drug resistance in many malignant tumour types. LncRNAs regulate the occurrence and development of many kinds of malignant tumours, such as rectal cancer 22 , gastric cancer 23 , lung cancer 24 , cervical cancer 25 and prostate cancer 26 .The lncRNA SBF2-AS1 is also involved in the formation of a variety of tumours [27][28][29] . Many lncRNAs have been demonstrated to be related to the occurrence of oesophageal squamous cell carcinoma, and many lncRNAs can cause a poor prognosis in patients with oesophageal squamous cell carcinoma 30 . In this study, we found that SBF2-AS1 was highly expressed in a variety of tumours and then analysed the TCGA database, which revealed that the expression level of SBF2-AS1 was significantly higher in oesophageal squamous cell carcinoma tissues than in normal tissues. Previous studies have shown that SBF2-AS1 can promote the development of gastric cancer through sponging miR-545 31 . Next, we wanted to determine whether SBF2-AS1 sponges other miRNAs to regulate the development of oesophageal squamous cell carcinoma.
In our study, we found networks of ceRNAs against SBF2-AS1, miR-338-3P and E2F1 in EC109 cells. Through bioinformatics analysis, we found that miR-338-3P and miR-362-3P may be downstream target genes of SBF2-AS1. The tumour suppressor genes miR-338-3P-32 32,33 and miR-362-3P 34,35 are downregulated in many tumours such as ESCC. In this study, RIP and RNA-pull down experiments showed that SBF2-AS1 could directly target miR-338-3P and miR-362-3P, which confirmed the results of bioinformatics analysis. In terms of an effect on cell proliferation, the overexpression of SBF2-AS1 alone could promote cell proliferation, while co-transfection with miR-338-3P and miR-362-3P abrogated the proliferation of tumour cells promoted by SBF2-AS1. These results indicate that SBF2-AS1 plays the role of a ceRNA against miR-338-3P and miR-362-3P. An increasing number of studies have shown that lncRNAs, miRNAs, and mRNAs form the core of the ceRNA network 36 . Through use of the TargetScan database, we identified a common target gene of miR-338-3P and miR-362-3P: EF21. Previous studies have shown that E2F1 is an oncogene that can promote the occurrence and development   37 . Some studies have shown that lncRNAs are involved in regulation of the miRNA/E2F1 axis in malignant tumours. For example, the lncRNA FER1L4 can downregulate the expression of miR-372 and enhance the expression of E2F1 in gliomas 38 . The lncRNA NNT-AS1 can promote the carcinogenesis and cell cycle progression of gastric cancer through the miR-424/E2F1 axis 39 . The results of this study showed that SBF2-AS1 enhanced expression of the E2F1 protein through acting as a sponge for miR-338-3P and miR-362-3P in ECA109 and TE-13 cells. The rescue experiment further showed that SBF2-AS1 promoted the growth of oesophageal squamous cell carcinoma cells through the miR-338-3P-miR-362-3P/E2F1 axis. Our study shows that the SBF2-AS1/miR-338-3P-miR362-3P/E2F1 axis is a potential mechanism for the malignant growth of oesophageal squamous cell carcinoma. Overall, our study confirmed the high expression of SBF2-AS1 in oesophageal squamous cell carcinoma. SBF2-AS1 promotes the occurrence and development of oesophageal squamous cell carcinoma through sponging miR-338-3P andmiR-362-3P, which regulates the expression of E2F1. This study suggests that SBF2-AS1 may be a molecular marker of the malignant proliferation of oesophageal squamous cell carcinoma.

Materials and methods
Tissue specimens. Specimens were collected from 150 patients at Taixing People's Hospital, and all specimens were frozen at -80 °C. With regard to the collection of specimens, all the patients signed an informed consent form, approval was obtained from the Ethics Committee of Taixing People's Hospital, and the experiment was carried out in accordance with the regulations of the Ethics Committee of Bengbu Medical College.
RNA extraction and qPCR. RNA was extracted with TRIzol according to the manufacturer's instructions (Life Technologies, Scotland, UK). The extracted RNA was reverse transcribed into cDNA using PrimeScript RT Master Mix (Takara, Catalogue No. RR036A). DNAMAN software was used to design primers according to the principle of primer design. qPCR was performed on a Bio-Rad CFX-96 system (Bio-Rad). The primers used are listed in the attached file.

Fluorescence in situ hybridization (FISH).
Through previous experiments, SBF2-AS1 was found to be mainly located in the cytoplasm, and detection of the subcellular localization of SBF2-AS1 was carried out with a Ribo Fluorescent in situ hybridization kit (RiboBio, Guangzhou, China, R11060.6). The cells were plated on a 24-wellplate, washed with PBS for 5 min after 24 h, and fixed with 4% paraformaldehyde for 10 min at room temperature. After washing with PBS three times, the cells were incubated with permeabilization buffer for 5 min at 4 °C. The cells were then sealed with pre-hybrid solution for 30 min at 37 °C, after which 2.5 µl of the SBF2-AS1 probe designed by Ribo was added to hybrid solution preheated to 37 °C. The pre-hybrid solution was discarded, and the hybrid solution containing probe was added and incubated overnight at 37 °C. Then, the cells were sequentially washed with hybrid solutions I, II and II and then stained with DAPI staining solution for 10 min, after which the cells were washed with PBS three times for fluorescence detection. Finally, a ZEISS microscope was used to acquire images under 40 × magnification.

RTCA .
Six-well plates were seeded with cells after they had grown to 80%-90% confluence. Twenty-four hours later, cell transfection was carried out. One day later, the baseline was measured by the addition of 50 µl of complete medium to each well of the E-plate. A total of 5 × 10 3 cells from each group were added to wells in the E-plate, and the data were then measured with an xCELLigence Real Time Cell Analysis System (ACEA Biosciences, 380601050).

Clone formation experiment.
A total of 400 cells in RPMI 1640 medium were added to each well of a six-well plate and then washed twice with pre-cooled PBS. The cells were then fixed with 4% paraformaldehyde at room temperature for 15 min and washed twice with PBS. Then, the cells were incubated for 20 min at room temperature with 1% crystal violet. Finally, tap water was used to slowly flush away the excess crystal violet.

5-Ethynyl-2′-deoxyuridine. The EdU assay was carried out with the Cell-Light EdU Apollo 567 In Vitro
Kit (RiboBio, Guangzhou, China, C10310-1). Cells in logarithmic growth phase were inoculated in a 96-well plate with 4 × 10 3 -1 × 10 4 cells per well, and the cells were transfected 24 h later. Twenty-four hours after transfection, EdU solution was added to the cells and incubated for 2 h, after which the cells were washed twice with PBS for 5 min each. A cell fixation solution (50 µl) was added and incubated for 30 min at room temperature, after which the fixation solution was discarded. Fifty microliters of glycine at 2 mg/ml was added to each well and incubated for 5 min to stain the cells. Then, the cells were washed for 5 min with pre-cooled PBS. Then, 100 µl of penetrant was added to each well for incubation for 10 min and washed for 5 min with PBS. Then, 100 µl of a Cell cycle analysis. Cells in a six-well plate were fixed with 500 µl of 75% ethanol at 4 °C for 4 h. Then, the cells were centrifuged at 1500 rpm for 5 min and washed once with PBS, and 400 µl of Piran solution was added to the cells. Then, the cells were incubated for 30 min with 100 µl of RNase at 4 °C avoiding light for flow cytometry detection.

Flow cytometry to detect apoptosis. An Annexin V-FITC Apoptosis Detection Kit (KeyGen Biotech,
Nanjing, China, KGA108) was used to detect cell apoptosis. Cells were inoculated on a six-well plate and washed twice with pre-cooled PBS at 24 h after transfection. The cells were digested with trypsin without EDTA, centrifuged at 300×g for 5 min at 4 °C, washed once with pre-cooled PBS, and centrifuged at 4 °C at 300×g for 5 min. Then, 250 µl of binding buffer was added to the cells to resuspend them, and 100 µl of cells was added to a flow tube. A PI solution (5 µl) and Annexin V-Alexa Fluor (10 µl) were mixed and reacted with the cells at room temperature for 15 min, after which flow cytometry analysis was performed.
Western blot analysis. The culture medium from the Petri dish containing the cells was discarded, and pre-cooled PBS was added to wash the cells three times for 5 min each. ESCA cells (ECA109 and TE-13) were resuspended in RIPA buffer and a protease inhibitor cocktail at a ratio of 100:1, and the ECA109 and TE-13 cells were lysed on ice for 30 min. The cells were scraped off of the dish and placed into a 1.5-ml EP tube. Then, the supernatant after centrifugation at 4 °C and 12,000×g for 15 min was obtained. Next, 5 × SDS buffer was added to the supernatant at a ratio of 4:1 according to the amount of supernatant, and the samples were placed in an incubator at 95 °C for 5 min. Finally, the protein concentration was determined with the BCA method. The volume of the solution containing 50 µg of protein was used as the sample volume. Samples were separated by SDS-PAGE at 70 V for 30 min and 110 V for 90 min. The proteins were transferred to a PVDF membrane by electroporation using an eBlot rapid protein transfer system. After transfer to the membrane, the membrane containing a band for the target protein was incubated with 1% skim milk (232,100) for 2 h. TBST was used to wash the membranes three times for 5 min each. The membranes were incubated with antibodies against E2F1 (Cell Signaling Technology, USA, 1:1000, 3742 s), Cyclind1 (Cell Signaling Technology, USA, 1:1000, 2978 T) and P21 (Cell Signaling Technology, USA, 1:1000, 2947 s) in a shaker at 4 °C overnight. After washing three times, the membranes were incubated with TBST containing goat anti-rabbit IgG and anti-horseradish peroxidase (HRP) while protected from light for 2 h. TBST was used to wash the membranes three times for 5 min each, and the membranes were finally scanned with an Odyssey fluorescence scanner.

RIP.
The Magna RIP RNA-Binding Protein Immunoprecipitation Kit (Millipore) was used to determine whether SBF2-AS1 and miR-362-3P/miR-338-3P can bind the AGO2 protein. The cells were washed twice with pre-cooled PBS for 5 min each, and the supernatant was discarded to collect the cells. The cells were lysed on ice with RIP lysis buffer for 5 min, and the cells were then frozen at − 80 °C. Fifty microlitres of magnetic beads was washed once with 500 µl of RIP wash buffer, placed on a magnetic stand, and the supernatant was discarded. Magnetic beads (50 µl) and antibodies (5 µg) were combined and incubated at room temperature for 30 min. The cell lysate was removed from − 80 °C and centrifuged at 14,000 rpm for 10 min at 4 °C. Then, 100 µl of cell lysate and RIP immunoprecipitation buffer containing magnetic beads were mixed and incubated overnight in a 4 °C incubator while shaking. The RNA from the magnetic bead-immunoprecipitation complexes was extracted by Figure 5. SBF2-AS1 in combination with miR-338-3P and miR-362-3P can increase the expression of E2F1 to promote the proliferation of oesophageal squamous cell carcinoma. (a) Western blot assay (samples derived from another experiment and the blots were processed in parallel). The SBF2-AS1 overexpression plasmid, a negative control, SI-SBF2-AS1 and SI-NC were transiently transfected into ECA109 and TE-13 cells, which were subjected to Western blot analysis of E2F1, Cyclind1 and P21 protein expression. (b) Western blot assay (samples derived for another experiment and the blots were processed in parallel). The SBF2-AS1 overexpression plasmid, a negative control, the SBF2-AS1 overexpression plasmid + miR-338-3P, and the SBF2-AS1 overexpression plasmid + miR-362-3P were transiently transfected into ECA109 and TE-13 cells, which were subjected to Western blot analysis of E2F1, Cyclind1 and P21 protein expression. (c) Colony formation assay. The SBF2-AS1 overexpression plasmid, a negative control, the SBF2-AS1 overexpression plasmid + miR-338-3P, and the SBF2-AS1 overexpression plasmid + miR-362-3P were transiently transfected into ECA109 and TE-13 cells, and colony formation ability was analysed by colony formation assay. (d) Analysis of cell proliferation. The SBF2-AS1 overexpression plasmid, a negative control, the SBF2-AS1 overexpression plasmid + miR-338-3P, and the SBF2-AS1 overexpression plasmid + miR-362-3P were transiently transfected into ECA109 and TE-13 cells, and cell proliferation was detected by EdU assay. (e) Analysis of cell proliferation. The SBF2-AS1 overexpression plasmid, a negative control, the SBF2-AS1 overexpression plasmid + miR-338-3P, and the SBF2-AS1 overexpression plasmid + miR-362-3P were transiently transfected into ECA109 and TE-13 cells, and the expression of SBF2-AS1 was analysed by RT-PCR. (f) Analysis of cell proliferation. The SBF2-AS1 overexpression plasmid, a negative control, the SBF2-AS1 overexpression plasmid + miR-338-3P, and the SBF2-AS1 overexpression plasmid + miR-362-3P were transiently transfected into ECA109 and TE-13 cells, and cell proliferation was detected by RTCA. Data are shown as the mean ± SEM. *P < 0.05, **P < 0.01, ***P < 0.01.

Establishment of xenograft tumours in nude mice.
A subcutaneous tumour-bearing model confirmed that SBF2-AS1 can promote the proliferation of oesophageal cancer in vivo. Seven-week-old male SPFgrade Balb/C nude mice were purchased from Yunqiao (Nanjing, China). The animal experiments described in the manuscript were approved by the Animal Research Ethics Committee of Bengbu Medical College. The study was carried out in compliance with the ARRIVE guidelines. ECA109 cells at logarithmic growth phase were collected and subcutaneously injected into nude mice at 5 × 10 5 cells per mouse. Measure the size and volume of the mouse tumor every 3 days. When the mouse tumor grows to 100 mm 3 , randomly divide 10 mice into two groups. SBF2-AS1ASO and SBF2-AS1ASO-NC (RiboBio, Guangzhou, China) were given to mice in the SBF2-AS1ASO and SBF2-AS1ASO-NC groups at 1.5 kg/ml twice a week for three weeks. The body weight and tumor www.nature.com/scientificreports/ size and volume of the mice were measured every 3 days. Finally, the mice were anesthetized and euthanized. The tumors of the mice were removed and the pictures were taken.
Immunohistochemistry. Paraffinized slices were dewaxed in water and incubated with 3% H 2 O 2 at room temperature for 10 min. The slices were rinsed with distilled water and soaked in PBS twice for 5 min each. Then, 5% normal goat serum was closed and incubated at room temperature for 10 min. The serum was removed, and antibody was added and incubated with the cells at 4 °C overnight. The slices were washed with PBS three times, and an appropriate amount of horseradish peroxidase-labelled streptavidin was added and incubated with the slices at 37 °C for 30 min, followed by development with a chromogenic agent for 15 min. Finally, the sections were rinsed with water, re-dyed, dehydrated, cleared and sealed.
Statistical analyses. SPSS 16.0 software was used for the statistical analyses in this study. This study used many statistical research methods. The Kaplan-Meier method was used to estimate survival probability, and one-way ANOVA or Student's t-test was used for intergroup comparisons. All data are shown as the average ± standard error. P < 0.05 was used to indicate statistical significance (Supplementary File S1).

Data availability
All data generated or analysed during this study are included in this published article (and its Supplementary Information files).