Long non‐coding RNA NR2F1‐AS1 promoted proliferation and migration yet suppressed apoptosis of thyroid cancer cells through regulating miRNA‐338‐3p/CCND1 axis

Abstract Thyroid cancer (TC) is a prevalent endocrine malignant cancer whose pathogenic mechanism remains unclear. The aim of the study was to investigate the roles of long non‐coding RNA (lncRNA) NR2F1‐AS1/miRNA‐338‐3P/CCND1 axis in TC progression. Differentially expressed lncRNAs and mRNAs in TC tissues were screened out and visualized by R program. Relative expression of NR2F1‐AS1, miRNA‐338‐3p and cyclin D1 (CCND1) was determined by quantitative real time polymerase chain reaction. In addition, Western blot analysis was adopted for evaluation of protein expression of CCND1. Targeted relationships between NR2F1‐AS1 and miRNA‐338‐3p, as well as miRNA‐338‐3p and CCND1 were predicted using bioinformatics analysis and validated by dual‐luciferase reporter gene assay. Besides, tumour xenograft assay was adopted for verification of the role of NR2F1‐AS1 in TC in vivo. NR2F1‐AS1 and CCND1 were overexpressed, whereas miRNA‐338‐3p was down‐regulated in TC tissues and cell lines. Down‐regulation of NR2F1‐AS1 and CCND1 suppressed proliferation and migration of TC cells yet greatly enhanced cell apoptotic rate. Silence of NR2F1‐AS1 significantly suppressed TC tumorigenesis in vivo. NR2F1‐AS1 sponged miRNA‐338‐3p to up‐regulate CCND1 expression to promote TC progression. Our study demonstrated that up‐regulation of NR2F1‐AS1 accelerated TC progression through regulating miRNA‐338‐3P/CCND1 axis.

more optimal therapies have remained to be further developed. 5,6 Hence, it is significant to unveil underlying mechanism of TC progression and pathogenesis, which may well provide us with more integrated and upgraded therapies for TC patients.
Long non-coding RNAs (lncRNAs) are non-coding polyadenylated RNAs typically longer than 200 base pairs and act as natural antisense transcripts in animals and humans. 7 Long non-coding RNA serves as one of vital epigenetics regulatory mechanism, as well as DNA methylation and genomic imprinting. 8 A great number of studies have revealed that lncRNAs can take part in massive biological processes, including transcriptional activation and interference, cellular differentiation, embryogenesis and stem cell biology. 9,10 Besides, dysfunction of lncRNAs has been clarified to be associated with progression of various diseases, such as lung cancer, liver cancer and breast cancer. [11][12][13] It was reported that some lncRNAs play pivotal roles in TC tumorigenesis, such as lncRNA CASC2 and lncRNA PROX1-AS1. 14,15 Furthermore, previous reports showed that lncRNA NR2F1-AS1 was able to regulate hepatocellular carcinoma oxaliplatin resistance by regulating miR-363/ABCC1 axis. 8 Nonetheless, it remains to be demonstrated whether and how NR2F1-AS1 functions in TC progression.
MicroRNAs are 18-23 nucleotide-long non-coding (ncRNAs) RNAs that always serve as negative regulators of translation and are involved in cell functions via binding to the 3′-untranslated region (UTR) of their targets. 16 In addition, dysregulation of miRNAs closely related to diseases, such as cancers, cardiovascular disease and diabetes. 17 In addition, studies revealed that some miRNAs, such as miR-520a-3p, miR-718 and miR-335-5p are closely related with progression of TC. [18][19][20] In addition, previous reports revealed that miR-338-3p participated in the progression of cervical cancer, colorectal cancer and renal cancer. [21][22][23] However, there are no sufficient evidences of roles of miR-338-3p, as well as the relationship between NR2F1-AS1 and miR-338-3p in the progression of TC.
Cyclin D1 (CCND1), one of the highly conserved members of the cyclin family, was well characterized by a periodicity in protein abundance in cell cycle. 24 Dysfunction of CCND1 was closely related with progression of several cancers by causing abnormal proliferation. 24 Several reports showed the targeted relationships between CCND1 and miRNAs in cancer. Liu et al showed that miR-138 inhibited nasopharyngeal carcinoma growth and tumorigenesis by targeting CCND1. Besides, Cai et al found that CCND1, CDK2 and CDK6 could be directly targeted by miR-186 in lung adenocarcinoma. 25 However, the roles of CCND1, as well as miR-338-3p/CCND1 axis in the TC tumorigenesis were little understood.
Competing endogenous RNA (ceRNA) mechanism proposed that transcripts such as mRNAs, pseudogenes and lncRNAs can serve as natural miRNA sponges by competitive binding to miRNA response elements to suppress their expression and function. 26 However, the ceRNA mechanism in TC still needs further study.
Our study is aimed to unveil the relationship between NR2F1-AS1 and miR-338-3p and their roles in TC. Our study revealed that dysregulation of NR2F1-AS1 and miR-338-3p affected the TC progression via targeting CCND1, which might well offer us with more potential therapeutic strategies for TC.

| Microarray analysis and co-expression network analysis
Microarray data were obtained from GEO (https ://www.ncbi.nlm. nih.gov/geo/query/ ). The series accession number of both lncRNA and mRNA expression profile was GSE3678 and matched platform was GPL570. In brief, data were processed for selecting differentially expressed both lncRNAs and mRNAs using selecting criteria adj P value <0.05 and |FC (fold change)| > 2 by limma package. Top differentially expressed ones were visualized by heatmap. Gene set enrichment analysis (GSEA) was made using GSEA v3.0 software and was adopted to perform gene set enrichment analysis. Co-expression network analysis was carried out based on the correlation coefficient which was calculated between differentially expressed lncRNAs and mRNAs. In brief, to establish lncRNA/mRNA co-expression network, 'pearson' in 'psych' package was employed for validating the correlations in the selected mRNAs and lncRNAs. The networks were then adjusted by 'BH'. Thereafter, they were graphed by Cytoscape software. In co-expression networks, nodes represented differently expressed genes, and the edges represented the co-expression status.
After finishing this construction, pENTR TM /U6-sh-NR2F1-AS1 was sequenced to verify the accuracy of the inserted sequence.
Thereafter, pENTR TM /U6-sh-NR2F1-AS1 was introduced into When cell confluence rate reached 80% to 90%, cell culture media should be replaced by serum-free fresh medium 3 hours before transfection.

| Protein isolation and Western blot analysis
Proteins were isolated by radio-immunoprecipitation assay buffer and protein concentration was determined using bicinchoninic acid (BCA) protein concentration assay Kit (Abcam). Thereafter, protein samples

| Cell Counting Kit-8 Assay
Cell proliferation ability was evaluated by Cell Counting Kit-8 (CCK-8; Engreen, Beijing, China). In brief, cells were placed in 96-well plates and pre-incubated with 15 μL CCK-8 solutions in each well. Optical density (OD) at 450 nm wavelength was measured every 24 hours after transfection.

| Cell migration assay
Cell migration rate was measured by Transwell chamber assay (

| Flow cytometry analysis
Cell apoptosis status was determined following protocol of Annexin V-FITC/PI-cell apoptosis Detection Kit (Biolegend, San Diego, CA).
After successively staining with both Annexin V-FITC and protease inhibitors for half an hour, cells were incubated in a dark room for 25 minutes. Cell apoptotic status was analysed using MoFlow flow cytometer in whole experimental process.

| RNA immunoprecipitation
RNA immunoprecipitation (RIP) assays were carried out using EZ-Magna RIP RNA-binding Protein Immunoprecipitation Kit (Millipore) following standard protocol. FTC-133 cells were lysed using RIP lysis buffer, and then cell lysates were incubated with RIP buffer, including magnetic beads conjugated to human anti-Ago2 antibody (Abcam), mouse IgG (Beyotime, China) served as control.
Co-precipitated RNAs were isolated using TRIzol reagent (TaKaRa, China). Ago-IPs in this study was detected at least three times.

| Tumour xenograft assay
All of the experimental protocols were permitted by Animal Care and

| Immunohistochemistry
Tissue sections were dried at 60°C for 1 hour and de-waxed by au- section will be exposed to a freshly prepared diaminobenzidine for 4-6 minutes and stained with hematoxylin for 15 seconds.

| Statistical analysis
Data were presented as mean value ± standard deviation (Mean ± SD) in this study. GraphPad Prism 6.0 (La Jolla, CA) was recruited to conduct statistical analysis. Multiple comparisons were conducted by the one-way ANOVA test. P value less than 0.05 was thought to be statistically significant.

| The expression of both NR2F1-AS1 and CCND1 was higher in TC tissues than normal tissues based on microarray analysis
It was reported that P53 signalling pathway played vital roles in tumour progression. Differentially expressed mRNAs in p53 pathway in TC tissues were selected out and visualized in Figure 1A. Among them, CCND1 showed a robust increase in TC tissues by contrast with adjacent normal tissues, which indicated its potential oncogenic role in TC. P53 signalling pathway was greatly up-regulated in TC as revealed in Figure 1B. Among differentially expressed lncRNAs, NR2F1-AS1 was significantly were higher in TC tissues in Figure 1C. According to co-expression network analysis, CCND1 was predicted to be co-expressed with NR2F1-AS1 ( Figure 1D). In short, the expression of both NR2F1-AS1 and CCND1 was higher in TC tissues than that in normal tissues based on microarray data.

| Knockdown of NR2F-AS1 suppressed cell proliferation and migration yet enhanced cell apoptosis in TC
After inspecting 25 TC tissues and paired adjacent normal tissues, NR2F-AS1 was greatly increased in TC tissues (Figure 2A; Besides, the result of qRT-PCR showed that NR2F-AS1 was up-regulated in FTC-133 cells and B-CPAP cells compared with Nthy-ori 3-1 ( Figure 2B; P < 0.01). qRT-PCR was adopted to detect the expression of NR2F-AS1 after transfection. As shown in Figure   S1, plasmid construction was successful. The results revealed that Taken together, down-regulation of NR2F-AS1 could inhibit cell viability and migration, whereas enhance cell apoptosis in TC.

| miR-338-3p shared a targeted relationship with NR2F1-AS1 and was down-regulated in TC
According to miRcode results, miR-338-3p was predicted to bind to the 3′-UTR of NR2F1-AS1-wt as shown in Figure 4A. Hence, dual-luciferase reporter assay was employed for further verification of the targeted relationship between miR-338-3p and NR2F1-AS1. It was demonstrated that relative luciferase activity in group co-transfected with miR-338-3p mimics and NR2F1-AS1-wt showed lower level than that of group with co-transfection of both NR2F1-AS1-wt and NC in Figure 4B, whereas there was no significant change in NR2F1-AS1mut group, revealing that there was the direct binding between miR-338-3p and NR2F1-AS1. As shown in Figure 4C, F I G U R E 5 A, Overexpression of miR-338-3p greatly suppressed proliferation, which was otherwise promoted by inhibition of miR-338-3p compared with negative control (NC) group; (B,C) Cell migration assay showed that overexpression of miR-338-3p greatly suppress migration, which was otherwise promoted by suppression of miR-338-3p compared with NC group. *P < 0.05, **P < 0.01, compared with NC group; # P < 0.05, ## P < 0.01, compared with inhibitor group

| NR2F1-AS1 promoted cell proliferation and cell migration yet suppressed cell apoptosis by downregulating miR-338-3p in TC
To unveil the roles of miR-338-3p/NR2F1-AS1 axis in TC, both CCK-8 assay and cell migration assay were adopted to assess cell proliferation rate and cell migration rate, respectively. It was unveiled by CCK-8 assay that increase in miR-338-3p could suppress cell proliferation, whereas inhibition of miR-338-3p could otherwise enhance cell proliferation in FTC-133 cells and B-CPAP cells ( Figure 5A;

| CCND1 was directly targeted by miR-338-3p and could be regulated by NR2F1-AS1/ miR-338-3p in TC
The results of bioinformatic analysis based on TargetScan revealed that CCND1 might be the target of miR-338-3p ( Figure 7A). To validate the targeted relationship between miR-338-3p and CCND1, dual-luciferase reporter gene assay was utilized. Relative luciferase activity in that group co-transfected with miR-338-3p mimics and CCND1-wt revealed lower level than that of group with co-transfection of both miR-338-3p NC and CCND1-wt, suggesting the direct binding between miR-338-3p and CCND1 in TC ( Figure 7B).
Afterwards, the relative mRNA expression level of CCND1 was greatly decreased CCND1 protein expression, which could be retrieved to normal levels co-transfected with miR-338-3p inhibitor.
In addition, Figure 8B revealed that sh-CCND1 significantly downregulated CCND1 protein expression, which was recovered to the normal expression level by co-transfection with miR-338-3p inhibitor in both cell lines. In short, CCND1 was targeted by miR-338-3p and could be regulated by NR2F1-AS1/miR-338-3p in TC.

| miR-338-3p suppressed proliferation and migration, whereas enhanced apoptosis of TC cells via down-regulation of CCND1
The results of CCK-8 assay revealed that inhibition of CCND1 greatly  Figure 11E.
In short, knockdown of NR2F1-AS1 inhibited TC tumorigenesis in vivo.

| D ISCUSS I ON
In our study, NR2F1-AS1 and CCND1 was greatly up-regulated, whereas miRNA-338-3p was otherwise suppressed in TC tissues and cell lines compared with adjacent normal tissues and cell lines.
Down-regulation of NR2F1-AS1 inhibited proliferation and migration, whereas enhanced apoptosis of TC cells by decreasing the expression of miRNA-338-3p yet increasing the expression of CCND1.
In brief, these findings will provide us with more significant contributions to further treatment of TC.
Long non-coding RNAs, as well as lncRNA/miRNA axis always played pivotal roles in the initiation and progression of a set of cancers. Previous report showed that lncRNA NEAT1 regulated TC F I G U R E 9 A, Down-regulation of cyclin D1 (CCND1) greatly suppressed cell proliferation, which was otherwise upregulated by miR-338-3p inhibitor in both FTC-133 cells and B-CPAP cells compared with negative control (NC) group; (B,C) Cell migration assay showed that knockdown of CCND1 greatly inhibited cell migration, which was instead promoted with miR-338-3p inhibitor in both FTC-133 cells and B-CPAP cells compared with NC group. **P < 0.01, compared with NC group; ## P < 0.01, compared with inhibitor group progression by modulating miR-129-5p expression. 27 Besides, ln-cRNA CCAT1 promoted cell growth and cell migration by down-regulation of miR-143 in TC cell line FTC-133. 28 Furthermore, lncRNA TUG1 influenced papillary thyroid cancer proliferation, migration and EMT formation through targeting miR-145. 29 In addition, NR2F1-AS1 regulated hepatocellular carcinoma resistance by regulating the expression of miR-363. 8 Our result revealed that the expression of NR2F1-AS1 was higher in TC tissues and cell lines than adjacent normal tissues and cell lines. In addition, NR2F1-AS1 promoted TC cell proliferation and migration yet suppressed cell apoptosis by downregulating miR-338-3p expression, which was in line with suppressor role of miR-338-3p in TC. 30 miRNAs always governed biological processes by repressing translation progress or triggering the degradation of its target mRNAs. 31 Furthermore, previous studies showed that miRNAs played pivotal roles in the treatment of tumours. 32 Our results showed that miR-338-3p was down-regulated in TC tissues and cell lines in comparison with adjacent normal tissues and cell lines, indicating that miR-338-3p might well function as tumour suppressor to affect TC progression. Several studies revealed the significance of miRNAs in TC. miR-214 regulated papillary thyroid carcinoma (PTC) cell metastasis and growth by regulating PSMD10. 33 In addition, miRNA-361-5p inhibited PTC progression by targeting ROCK1. However, the roles of miR-338-3p in TC progression, especially the relationship with mRNA were still poorly understood. miR-338-3p was reported to function as a tumour suppressor to inhibit renal cell carcinoma progression. 34 miR-338-3p suppressed growth and invasion of non-small cell lung cancer cells by targeting IRSA2. 35 Besides, miR-338-3p inhibited TC progression through targeting AKT3. 30 Our findings revealed that miR-338-3p suppressed cell proliferation and migration, whereas enhanced cell apoptosis in TC by targeting CCND1, which is similar with the roles of miR-338-3p/CCND1 axis in HCC. 36 Cyclin D1, a member of the family of cyclins, can control the passage of proliferating cells through cell cycle. 37 Up-regulation of CCND1 was observed in both benign and malignant thyroid tumours. 37 Dys-regulation of cyclin is closely related with oncogenesis and CCND1 functions as an oncogene in tumour progression. It was reported that CCND1 could be directly targeted by miR-186 to affect lung adenocarcinoma progression. 25 In addition, previous report showed that CCND1 could be directly targeted by miR-195 to affect tumour growth and metastasis in PTC Cell Lines. 24 Our results revealed that CCND1 was greatly up-regulated in TC tissues and cell lines. Besides, knock-down of CCND1 greatly suppressed cell proliferation and migration yet promoted cell apoptosis in TC, which was significant for the validation of the oncogenic roles of CCND1 in TC.
Besides, CCND1 may well be selected as a potential therapeutic target for the TC.
However, the limitations of this study should not be overlooked. To begin with, the study population was not big enough for further validation of the great significance of NR2F1-AS1/miR-NA-338-3p/CCND1 axis in the progression of TC. Besides, CCND1 may be down-stream of other miRNAs apart from miR-338-3p in TC. Similarly, miR-338-3p might well target other genes besides CCND1 in TC. Furthermore, diagnosis and treatment of TC is far F I G U R E 1 0 (A,B) Flow cytometry assay showed that down-regulation of cyclin D1 (CCND1) significantly increased apoptotic rate, which could be otherwise decreased with miR-338-3p inhibitor in both FTC-133 cells and B-CPAP cells in comparison with the negative control (NC) group. **P < 0.01, compared with NC group; ## P < 0.01, compared with inhibitor group more sophisticated than our expectations. Therefore, it really does matter to identify new lncRNAs, miRNAs and mRNAs, as well as interpret the mechanisms to facilitate therapeutic development against TC.
Above all, NR2F1-AS1 acted as a ceRNA of miR-338-3p and releasing CCND1 to promote the development of TC, which might well aid intervention strategies of TC in the future.

E TH I C S A PPROVA L A N D CO N S E NT TO PA RTI CI PATE
The procedures in the study were scrutinized and approved by

ACK N OWLED G EM ENTS
None.

CO N FLI C T O F I NTE R E S T S
The authors declare that they have no conflicts of interest with the contents of this article.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.