Long non‐coding RNA HOTAIR/microRNA‐206 sponge regulates STC2 and further influences cell biological functions in head and neck squamous cell carcinoma

Abstract Objective It is essential to characterize underlying molecular mechanism associated with head and neck squamous cell carcinoma (HNSCC) and identify promising therapeutic targets. Herein, we explored role of homeobox transcript antisense RNA (HOTAIR) in HNSCC to regulate stanniocalcin‐2 (STC2) by sponging microRNA‐206 (miR‐206). Methods HNSCC‐related differentially expressed genes and regulation network amongst HOTAIR, miR‐206 and STC2 were identified. Next, effect of HOTAIR on cell biological functions of HNSCC was identified after transfection of cells with HOTAIR overexpressed plasmids or siRNA against HOTAIR. PI3K/AKT signalling pathway‐related gene expression was measured after miR‐206 and STC2 were suppressed. Cell invasion, migration and proliferation were assessed. Finally, tumour growth was assessed to determine the effects of HOTAIR/miR‐206/STC2 axis in vivo. Results HOTAIR specifically bound to miR‐206 and miR‐206 targeted STC2. Downregulated HOTAIR or upregulated miR‐206 suppressed HNSCC cell proliferation, invasion and migration. miR‐206 inhibited PI3K/AKT signalling pathway by down‐regulating STC2. Besides, silenced HOTAIR or overexpressed miR‐206 repressed the tumour growth of nude mice with HNSCC. Conclusion HOTAIR regulated HNSCC cell biological functions by binding to miR‐206 through STC2.

Stanniocalcin-2 (STC2) involves in calcium and phosphate homoeostasis. 7 Besides, STC2 expression has been widely determined in metastatic cancers with a vital role played in metastasis and progression of lung cancer. 8 Interestingly, STC2 has been verified as target of microRNA-184 (miR-184), and miR-184 could suppress STC2 impeding the proliferation, invasiveness and migration capacity of glioblastoma cells. 9 Evidences have been presented supporting the functions of microRNAs (miRNAs) in regulating the metastatic process of tumours in various cancers, serving as oncogenes or tumour suppressors. 10 Moreover, the tumour-suppressing role of miR-206 has been demonstrated in several malignancies, such as ovarian cancer. 11 Notably, accumulating evidence has delineated roles of long non-coding RNAs (lncRNAs) in cancer initiation and progression by regulating miRNAs. 12 Especially, lncRNA homeobox transcript antisense RNA (HOTAIR) sponges miR-331-3p to promote tumorigenesis of gastric cancer. 13 In addition, lncRNAs, as tumour suppressors or oncogenes, involves in fundamental biological processes like cell proliferation, apoptosis and tumorigenesis. 14,15 Furthermore, lncRNAs are implicated in the carcinogenesis and progression of HNSCC. 16 Previous studies have identified upregulated HOTAIR expression in multiple cancer types, such as colorectal cancers and gastric cancer. 17,18 Meanwhile, a correlation between HOTAIR overexpression and breast cancer metastasis and prognosis as well as hepatocellular carcinoma has been delineated in existing researches. 19,20 These findings led to a hypothesis that HOTAIR participated in development of HNSCC by serving as a ceRNA of miR-206.

| Cell treatment
The HNSCC cell lines Tu686, TSCCA and Cal27 were purchased from American Type Culture Collection (Manassas, VA, USA). The

| Fluorescence in situ hybridization (FISH)
Ribo TM lncRNA FISH probe Mix (Red) and Ribo TM miR-206 FISH probe Mix (Green) (Guangzhou RIBOBIO Co., Ltd., Guangzhou, Guangdong, China) were performed. Briefly, the HNSCC cells were inoculated into cover slips placed in a 6-well plate for 1 day, fixed, then treated with proteinase K (2 μg/mL), glycine and acetamidine reagent, and finally added with 250 μL pre-hybrid solution at 42°C for 1 hour. Cells were added with 250 μL hybrid solution containing the probe (300 ng/mL) at 42°C overnight. After 3 washes with PBSTween (PBST), cells were added with 4',6-diamidino-2-phenylindole staining solution for nuclear staining and then added into a 24-well plate and incubated for 5 minutes. After that, cells were sealed with anti-fluorescent quencher. Five different fields were observed and photographed under a fluorescence microscope (Olympus Optical Co., Ltd., Tokyo, Japan).

| RNA-Binding protein immunoprecipitation (RIP)
HNSCC cells were lysed by radioimmunoprecipitation assay for 5 minutes. One proportion of cell extract served as input, and the remaining incubated with antibody and magnetic beads for binding. Magnetic bead-antibody complex was added with RIP wash buffer and lysed cell at 4°C overnight. Thereafter, magnetic bead-protein complex was obtained. Sample and input were treated with proteinase K separately to obtain RNA for PCR. argonaute (AGO) (ab32381, 1:50, Abcam Inc, Cambridge, MA, USA). Immunoglobulin (IgG; ab109489, 1:100, Abcam Inc, Cambridge, MA, USA) was served as NC.

| RNA isolation and quantitation
Total RNA was obtained. Primers (Table 1)

| Western blot analysis
Head and neck squamous cell carcinoma tissues were ground into a homogenate and added with lysis buffer to isolate total protein.

| Transwell assay
Cells were seeded in a 6-well plate for 48 hours, resuspended in serum-free DMEM, with cell density to 3 × 10 5 cells/mL. Then, 100 μL cell suspension was added to apical chamber and 500 μL DMEM containing 10% FBS to basolateral chamber in a 37°C incubator with 5% CO 2. After 24 hours, chamber was added with methanol for 10 minutes and crystal violet for 10 minutes. Cells in apical chamber were sealed with a neutral resin in a coverslip. Six randomly selected fields were observed with a microscope (100×) for counting.

| Scratch test
Cells (1 × 10 6 cells/mL) were added into 6-well plate till grew on the plate and then starved for 12 hours. Lines were drawn vertically on the plate, and the scratched cells were washed off. Photographs were taken at 0 and 48 hours. Fifteen lines were evenly distributed in scratch photograph. The width of scratches across the lines was measured.

| EdU staining
The cells were cultured with 50 μmol/L EdU (EdU labelling kit, Ribobio, Guangzhou, China) for 12 hours, fixed with polyformaldehyde and cultured in 5% glycine for 5 minutes. After being treated with 0.5% Triton X-100, the cells were added with Triton X-100, then anti-EdU antibody and then stained with Hoechst 33342. FITC and PI fluorescence were detected at 488 nm.

| Xenograft tumour in nude mice
Athymic female nude mice (4-6 weeks) from Animal Experimental Center of Peking University First Hospital (Beijing, China) were raised at (25 ~ 27°C) and constant humidity (45% ~ 50%). The HNSCC cells (1 × 10 7 cells/mL) were made into cell suspension and then inoculated into axilla of nude mice. The tumour volume (TV) was calculated as TV = 0.5 × a × b 2 (a, longest diameter, b, shortest diameter of tumour), and growth curve was drawn. After 5 weeks, mice were euthanized with tumours weighed.

| Statistical analysis
Data were processed using SPSS 21.0 statistical software (IBM Corp., Armonk, NY, USA), with normality and homogeneity tested.
F I G U R E 1 HOTAIR is upregulated in HNSCC cells. A, the expression of STC2 in HNSCC tissues and normal tissues in TCGA database; B, the survival of patients with different expression of STC2; C, the intersection of miRNA targeting STC2 predicted by four prediction websites; D, the expression of HOTAIR in HNSCC tissues and normal tissues; E, the expression of HOTAIR in five HNSCC cell lines and normal oral epithelial cell line detected by RT-qPCR; One-way ANOVA was used for statistical analysis, and the experiment was performed in triplicates; data were represented as mean ± standard deviation; *P < 0.05 vs HIOEC cell; STC2, stanniocalcin-2; HNSCC, head and neck squamous cell carcinoma; RT-qPCR, reverse transcription quantitative polymerase chain reaction; HOTAIR, homeobox transcript antisense RNA; ANOVA, analysis of variance The data with normal distribution were presented as mean ± standard deviation. The independent sample t test was employed for statistical analysis between two groups, and the comparison amongst groups was analysed by one-way analysis of variance (ANOVA).
The pairwise comparison amongst multiple groups was performed by Tukey's post hoc test. Besides, the TV at different time points F I G U R E 2 HOTAIR silencing suppresses the proliferation, migration and invasion while elevating apoptosis of HNSCC cells. A-D, the migration and invasion of HNSCC cells treated with HOTAIR overexpressed plasmids or siRNA against HOTAIR detected by transwell assay (×100); E-F, the cell migration and scratch healing rate HNSCC cells by scratch test; G, cell growth curve detected by CCK-8; H-I, proliferation and fluorescence intensity assayed by EdU; J-K, cell cycle and cell cycle ratio detected by flow cytometry; L-M, cell apoptosis assayed by flow cytometry. One-way ANOVA was used for statistical analysis amongst groups, and the experiment was performed in triplicates; data were represented as mean ± standard deviation; *P < 0.05 vs the blank or NC group

| High HOTAIR expression in HNSCC cells
DEGs related to HNSCC from TCGA database revealed that STC2 was highly expressed in HNSCC and correlated to HNSCC prognosis ( Figure 1A-B). Various online sources verified that miR-206 targeted STC2 ( Figure 1C) (miRDB, Starbase website, miRSearch website and the Mirtarbase website). HOTAIR was highly expressed in HNSCC ( Figure 1D). RT-qPCR demonstrated that HOTAIR expression in HNSCC cells was higher than that in HIOEC cells (P < 0.05). HOTAIR was differentially expressed in 5 HNSCC cell lines ( Figure 1E). In this experiment, Tu686 cells with the highest HOTAIR expression were selected.

| HOTAIR silencing inhibits HNSCC cell biological functions
Transwell results showed that compared with blank and NC, HNSCC

| Subcellular localization of HOTAIR
Through the analysis on http://lncat las.crg.eu/, HOTAIR was expressed in both nucleus and cytoplasm ( Figure 3A). Furthermore, FISH assay showed that in the Tu686 cells, HOTAIR was expressed in both the nucleus and cytoplasm. The green part depicted HOTAIR expression, and the blue part represented the nucleus ( Figure 3B).

| HOTAIR competitively binds to miR-206
The binding of HOTAIR and miR-206 was predicted through bioinformatics website RNA22 ( Figure 4A). Dual luciferase reporter assay ( Figure 4B) revealed that in contrast to NC, luciferase activity in the miR-206-Wt had decreased (P < 0.05), but that of miR-206-Mut almost remained the same (P > 0.05). RT-qPCR ( Figure 4C) revealed that in HNSCC cells following HOTAIR treatment HOTAIR expression was increased, miR-206 expression was decreased, while HNSCC cells treated with si-HOTAIR showed opposite results, suggesting that HOTAIR competitively bound to miR-206 and inhibited its expression. RIP ( Figure 4D) showed that HOTAIR could bind to AGO protein. Dual luciferase reporter assay showed that HOTAIR could co-localize with miR-206 in cytoplasm ( Figure 4E).

| miR-206 directly targets STC2
The analysis from bioinformatics website microRNA.org revealed the presence of a specific binding region between 3'UTR of STC2

| Overexpressed miR-206 or silenced STC2 suppresses HNSCC cell biological function via PI3K/ AKT signalling pathway
To further assess the effect of miR-206/STC2/PI3K/Akt signalling pathway on the biological function of HNSCC cells, transwell assay was firstly used to detect HNSCC cell migration ( Figure 7A Figure 7E-F).
F I G U R E 5 miR-206 targets STC2. A, verification of the targeting relationship between miR-206 and STC2; B, the luciferase activity of STC2-Wt and STC2-Mut detected by dual luciferase reporter assay; C, the expression of STC2 and miR-206 assessed by RT-qPCR; D, the grey value of STC2 protein band in response to the treatment of miR-206 mimic or miR-206 inhibitor; E, the protein level of STC2 determined by Western blot analysis; one-way ANOVA was used for different factors analysis, and the independent sample t test was used for statistical analysis between two groups; the experiment was performed in triplicates; data were represented as mean ± standard deviation; *, P < 0.05 vs the NC group or the blank group. Wt; wide type; Mut, mutant  Therefore, STC2 silencing could reverse the biological function changes of Tu686 cells induced by HOTAIR.

| HOTAIR silencing or miR-206 upregulation reduces tumour growth in nude mice
The results ( Figure 9A-

| D ISCUSS I ON
HNSCC is characterized by high recurrence, metastasis and unsatisfactory treatment results. 21 There is evidence indicating that the dysregulation of lncRNAs functions crucially in the genesis and development of HNSCC. 22 For instance, HOTAIR exerts a prognostic effect over biological functions of HNSCC. 23 Notably, HOTAIR has been demonstrated to be correlated with cell apoptosis and proliferation in various human malignancies, including HNSCC. 24  A, the tumour volume of nude mice measured by a vernier caliper; B, xenograft tumours in nude mice observed by naked eyes; C, tumour weight of nude mice; one-way ANOVA was used for analysis amongst groups, and the experiment was performed in triplicates; data were represented as mean ± standard deviation; *P < 0.05 vs the blank or NC group   32 Notably, STC2 upregulation increased HNSCC cell proliferation, invasion and migration, tumour growth, and metastasis, revealing that STC2 could be a novel strategy for HNSCC treatment. 33 Further, STC2 is a target of miR-206, and miR-206 could downregulate STC2 expression. In consistency with our results, miR-206 inhibited tumour growth and metastasis in GC via targeting STC2. 34 The impact of activated PI3K/AKT signalling pathway is significant in various fundamental biological activities. 35 Likewise, the PI3K/AKT signalling pathway regulated cell biological functions in HNSCC. 36 Furthermore, overexpressed miR-206 suppressed lung cancer cell migration and invasion via inhibition of the PI3K/AKT/ mTOR signalling pathway. 37 In conclusion, silencing HOTAIR could inhibit HNSCC biological functions via STC2 downregulation by competitively binding to miR-206. HOTAIR could competitively bind to miR-206, thereby stimulating STC2 expression, activating PI3K/AKT signalling pathway ( Figure 10). Thus, HOTAIR silencing can serve as a therapeutic target for HNSCC. However, further studies with larger sample sizes are needed to elucidate specific mechanisms of HOTAIR in HNSCC.

ACK N OWLED G EM ENT
The authors want to show their appreciation to reviewers for their helpful comments.

CO N FLI C T O F I NTE R E S T
None.

AUTH O R ' S CO NTR I B UTI O N
TCL and YQ designed the study. ZZ collated the data. Erik Schiferle and SFX carried out data analyses and produced the initial draft of the manuscript. HS and TCC contributed to drafting and polishing the manuscript. All authors have read and approved the final submitted manuscript.