Knockdown of circ_0011946 targets miR-216a-5p/BCL2L2 axis to regulate proliferation, migration, invasion and apoptosis of oral squamous cell carcinoma cells

Background Circular RNAs (circRNAs) are implicated in the development of oral squamous cell carcinoma (OSCC). The aim of current research is to elucidate the role and mechanism of circ_0011946 in the functional behaviors of OSCC cells. Methods Circ_0011946, microRNA (miR)-216a-5p, B cell lymphoma-2-like 2 protein (BCL2L2) abundances were exposed by quantitative reverse transcription polymerase chain reaction (qRT-PCR) or western blot. Cell proliferation, migration, invasion and apoptosis were detected by MTT, colony formation assay, transwell, wound-healing and flow cytometry assays, respectively. Target correlation was tested by dual-luciferase reporter and RNA pull-down assays. An in vivo xenograft experiment was employed to investigate the function of circ_0011946 on tumor growth in vivo. Results Circ_0011946 and BCL2L2 levels were increased, while miR-216a-5p level was decreased in OSCC tissues and cells. Circ_0011946 knockdown impeded proliferation, migration, and invasion, but promoted apoptosis in OSCC cells. Circ_0011946 functioned as a sponge for miR-216a-5p, and BCL2L2 was targeted by miR-216a-5p. Besides, miR-216a-5p or BCL2L2 knockdown partly attenuated the inhibitory influences of circ_0011946 silence or miR-216a-5p overexpression on OSCC cell progression. Furthermore, circ_0011946 post-transcriptionally regulated BCL2L2 through sponging miR-216a-5p. Moreover, circ_0011946 knockdown constrained OSCC tumor growth in vivo. Conclusion Circ_0011946 silence repressed OSCC cell proliferation, migration, and invasion, but promoted apoptosis through the regulation of the miR-216a-5p/BCL2L2 axis. Supplementary Information The online version contains supplementary material available at 10.1186/s12885-021-08779-4.


Background
Oral cancer is a common oral disease, and the major risk factors are tobacco use and alcohol consumption [1]. Oral squamous cell carcinoma (OSCC) is one of the primary cases of oral cancer, with low survival rate and high incidence [2]. Exploring the pathogenesis of OSCC may support a new therapeutic strategy for OSCC.
Noncoding RNAs are related to OSCC progression [3]. Circular RNAs (circRNAs) with covalently closed structure [4], which have been implicated in the development of various cancers, such as OSCC [5,6]. For instance, circ_0000140 can inhibit cell proliferation, metastasis by mediating miR-31/Hippo pathway in OSCC [7]. Circ_ 002178 can promote cell proliferation and migration by activating the protein kinase B/mammalian target of rapamycin (AKT/mTOR) pathway in OSCC [8]. Cir-cRNA hsa_circ_0011946 (circ_0011946), derived from the back-splicing of exons of Scm polycomb group protein homolog 1 (SCMH1), is capacity of promoting OSCC cell development by regulating proliferating cells nuclear antigen (PCNA) [9]. Nevertheless, our understanding of the mechanisms of circ_0011946 regulation in OSCC is still limited.
MicroRNAs (miRNAs) are another type of small noncoding RNA molecules, and dysregulated miRNAs are associated with OSCC progression [10,11]. A previous study uncovered that miR-216a-5p can repress the malignant phenotypes of OSCC via regulating eukaryotic translation initiation factor 4B (EIF4B) [12]. B cell lymphoma-2-like 2 protein (BCL2L2) belongs to the BCL2 family that exerts an important role in human cancers [13]. Moreover, BCL2L2 constrained the tumorigenesis of human cancers, like hepatocellular carcinoma and non-small cell lung cancer [14,15]. More importantly, BCL2L2 can promote tumor development by increasing cell proliferation in OSCC [16]. However, whether miR-216a-5p and BCL2L2 participate in the regulation of circ_0011946 in OSCC development is undiscovered.
In present research, we scrutinized the influences and mechanism of circ_0011946 in the malignant phenotypes of OSCC cells.

Tissues collection
OSCC tumor tissues and paired adjacent normal samples (> 2 cm from tumor tissue) were collected 30 OSCC patients, who undergo surgery between 2012 and 2015 at Jiaxing Hospital of Traditional Chinese Medicine, Jiaxing University. Patients received any other therapy before surgery were excluded. All tissues were stored in liquid nitrogen. The clinical-parameters of OSCC patients were provided in Table 1. The research got the permission of the ethics committee of Jiaxing Hospital of Traditional Chinese Medicine, Jiaxing University with the obtained written informed consent.

Quantitative reverse transcription PCR (qRT-PCR)
RNA was lysed using Trizol (Applygen, Beijing, China). Nuclear and cytoplasmic RNA were prepared via the Cytoplasmic & Nuclear RNA Purification Kit based on the manufacturer's suggestion (Norgen Biotek, Thorold, ON, Canada). 1 μg RNA was exposed to reverse transcription with specific Reverse Transcriptase kit (iGene-Bio, Guangzhou, China). The synthesized cDNA with SYBR (Toyobo, Osaka, Japan) and primers (Table 3) (Genscript, Nanjing, China) was used for qRT-PCR assay. The primer sequences were displayed in Table 2. U6 or GAPDH served as an internal control. Relative RNA level was decided by the 2 -ΔΔCt method [17], with U6 or GAPDH as an interior control.

Actinomycin D and RNase R treatment
For the stability of circ_0011946, OSCC cells were stimulated with Actinomycin D (

Colony formation assay
Briefly, 200 cells were inoculated into the 6-well plates for 10 days, and were spotted with 0.1% crystal violet (Solarbio, Beijing, China). The images were photographed, and the number of colonies (> 50 cells) was computed.

Transwell and wound healing analyses
Cell invasion was analyzed with Matrigel-coated transwell chambers (Costar, Corning, NY, USA). Cell migration was tested using transwell chambers without Matrigel and wound healing assay. For transwell assay, CAL27 and SCC25 cells (1 × 10 4 for migration assay, 3 × 10 4 for invasion analysis) in serum-free DMEM were plated into the superior chambers, whereas the lower chamber was supplemented with 500 μL DMEM with 10% serum. 24 h upon incubation, migrated or invasive cells were spotted with 0.1% crystal violet, and examined under a microscope (× 100 magnification, Nikon, Tokyo, Japan) with at least 6 random fields. For wound healing assay, 2 × 10 5 CAL27 and SCC25 cells after various transfections were placed into each well of 6-well plates until 80% of confluency was reached. Next, a "wound" was made using a 200 μL pipette tip, and cells were allowed to migrate for additional 24 h. The images (× 100 magnification) at 0 and 24 h were recorded. Relative migratory rate was calculated by the formula: (si-circ_0011946 (wound area)0h )-(si-circ_ 0011946 (wound area)24h )/(si-NC (wound area)0h )-(si-NC (wound area)24h ).

Xenograft experiment
BALB/c nude mice (male, 5-week-old) were procured from Beijing Laboratory Animal Center (Beijing, China) and fed in specific pathogen-free microisolator cages. Lentiviruses expressing shRNA-circ_0011946 (sh-circ_ 0011946) and the negative control (sh-NC) were supplied by Ribobio (Guangzhou, China). Stable CAL27 cells were established by transducing with sh-circ_ 0011946 or sh-NC, and the selection of puromycin. Stable CAL27 cells (3 × 10 6 /mouse) were hypodermically inoculated into the light flanks of the mice (n = 6/group). Tumor volume was estimated every week based on the formula: length × width 2 /2. 4 weeks later, all mice were euthanized using 5% isoflurane. Tumor tissues were collected and weighed. circ_0011946, miR-216a-5p and BCL2L2 levels in tumor samples were measured. Proliferation of the tumors was assessed with paraffinembedded tumor tissues by immunohistochemistry under the standard method using the anti-Ki67 antibody (ab15580, 1:100 dilution, Abcam), biotinylated goat antirabbit IgG secondary antibody (ab64256, 1:300 dilution, Abcam) and a 3,3-diaminobenzidine (DAB) Kit (Vector Laboratories, Peterborough, UK). The animal experiments got the approval of the Institutional Animal Care and Use Committee of Jiaxing Hospital of Traditional Chinese Medicine, Jiaxing University.

Statistical analysis
Data of three repetitions were represented as mean ± standard deviation (SD) through GraphPad Prism 7 (GraphPad Inc., La Jolla, CA, USA). Student t-test or ANOVA followed by Tukey test was exploited to analyze the difference. The Pearson's rank correlation coefficient was exploited to ascertain the correlation of genes in OSCC tissues. P < 0.05 was deemed as statistical significant.

Circ_0011946 abundance is elevated in OSCC
To analyze whether circ_0011946 was associated with OSCC development, circ_0011946 expression was detected in OSCC tissues and cells. Circ_0011946 level was higher in OSCC tissues compared with normal specimens (Fig. 1A). Moreover, circ_0011946 expression was apparently upraised in CAL27 and SCC25 cells in comparison to HOK cells (Fig. 1B). Furthermore, the ringlike structure of circ_0011946 was certified, since circ_ 0011946 was resistant to Actinomycin D and RNase R digestion than linear GAPDH and linear SCMH1 ( Fig.  1C and D). Additionally, subcellular localization assays showed that circ_0011946 predominantly localized to the cytoplasm (Fig. 1E). Thus, circ_0011946 might contribute to OSCC development.

MiR-216a-5p restrains OSCC cell development by managing BCL2L2
To test the impacts of miR-216a-5p and BCL2L2 in the malignant phenotypes of OSCC, gain-of-function assays were fulfilled in CAL27 and SCC25 cells. Elevated expression of miR-216a-5p obviously constrained cell proliferation, which was reversed by BCL2L2 restoration (Fig. 6A and B). Additionally, miR-216a-5p overexpression impelled a notable inhibition on cell migratory and invasive abilities, and this effect was mitigated by BCL2L2 up-regulation (Fig. 6C-E). Moreover, miR-216a-5p overexpression notably fostered cell apoptosis, which was attenuated by introduction of BCL2L2 expressing plasmid (Fig. 6F). These data uncovered that miR-216a-5p impacted OSCC cell behaviors by regulating BCL2L2.

Circ_0011946 knockdown reduces tumor growth in xenograft model
Subsequently, an in vivo xenograft model was utilized to examine the influence of circ_0011946 in OSCC cell growth. As displayed in Fig. 7A and B, tumor volume and weight of sh-circ_0011946 group were diminished than those in sh-NC group. Besides, circ_0011946 and BCL2L2 levels were lessened in sh-circ_0011946 group as opposed to sh-NC group, while miR-216a-5p level was increased (Fig. 7C-E). Additionally, sh-circ_ 0011946-transduced tumors had significantly fewer cells stained for Ki67 staining than the sh-NC control (Fig.  7E). These data showed that silencing of circ_0011946 minimized tumor growth in vivo.

Discussion
OSCC is the commonest type of oral cancers with high morbidity worldwide [18]. Many circRNAs have been revealed to apply important works in tumorigenesis and treatment of OSCC [6]. The objective of this study is to study new mechanism underlying the regulation of circ_ 0011946 in OSCC.
Meng et al. disclosed that circ_0011946 expression was strengthened in OSCC, and circ_0011946 silence restrained cell proliferation and metastasis in CAL27 cells [9]. Here, circ_0011946 abundance was elevated in OSCC. Moreover, circ_0011946 knockdown could Fig. 6 The influence of miR-216a-5p and BCL2L2 on OSCC progression. Cell viability (A), colony formation (B), migration and invasion (C-E), and apoptosis (F) were examined in CAL27 and SCC25 cells with transfection of miRNA NC, miR-216a-5p mimic, miR-216a-5p mimic + pc-NC or pc-BCL2L2. * P < 0.05 suppress cell proliferation, migration, and invasion, but promote apoptosis in CAL27 and SCC25 cells. This was also in agreement with that in other cancers, such as breast cancer and hepatocellular carcinoma [19,20]. Additionally, in this study, we used siRNAs targeting circ_0011946 to silence circ_0011946, which are designed by targeting the sequence spanning the junction of circ_0011946. Since the sequence spanning the junction is unique for circ_0011946, the siRNAs do not affect the transcription of the linear mRNA.
Next, we confirmed that BCL2L2 was a target and functional target of miR-216a-5p. BCL2L2 has established roles in promoting human tumorigenesis [14,[25][26][27]. Previous work also reported the oncogenic role of BCL2L2 in OSCC [16]. Moreover, BCL2L2 has not only anti-apoptotic role but also promotes cell migration and invasion [13]. Multiple studies have reported that BCL2L2 enhanced cell migration and invasion in several forms of cancers [15,28,29]. Here, we first uncovered that circ_0011946 regulated BCL2L2 level by miR-216a-5p competition. In addition, a xenograft model identified the anti-growth function of circ_0011946 silence in OSCC. However, there was a limitation that we did not analyze the influence of miR-216a-5p/BCL2L2 axis on xenograft tumor growth in vivo, which will be studied in future work.

Conclusions
Taken together, our findings suggest that circ_0011946 knockdown constrains cell proliferation, invasion and migration, but promotes apoptosis in OSCC at least in part by the miR-216a-5p/BCL2L2 axis.
Additional file 1: Supplement Figure 1. Expression correlation between BCL2L2 mRNA and miR-216a-5p levels in OSCC tissues using the Pearson's rank correlation coefficient.