Hsa_circ_0060937 accelerates non-small cell lung cancer progression via modulating miR-195-5p/HMGB3 pathway

ABSTRACT

Circular RNAs (circRNAs) exert a critical effect on tumorigenesis and development. Our research aimed to clarify the function and underlying mechanism of circ_0060937 inNSCLC. The concentrations of circ_0060937, miR-195-5p and high-mobility group box 3 (HMGB3) were monitored via qRT-PCR and western blot assays. Additionally, cell proliferation, apoptosis, migration and invasion were assessed using CCK-8, colony formation, flow cytometry and transwell assays. Glycolysis was evaluated via detecting glucose uptake and lactate product. The association between miR-195-5p and circ_0060937/HMGB3 were validated using dual-luciferase reporter, RNA pull-down and RIP assays. Furthermore,in vivo experiment was performed to analyze tumorigenesis.Circ_0060937 and HMGB3 levels were elevated, whereas miR-195-5p level was dropped in NSCLC. Circ_0060937 down-regulation restrainedNSCLC cell proliferation, migration, invasion and glycolysis, and triggered apoptosis. Knockdown of circ_0060937 restrained NSCLC development via absorbing miR-195-5p. Circ_0060937 silencing inhibited NSCLC progression by mediating HMGB3. Besides, circ_0060937 depletion suppressed tumor growth in vivo.Circ_0060937 knockdown hindered NSCLC development and glycolysis via regulating miR-195-5p/HMGB3 pathway.

Introduction

According to global cancer statistics in 2018, lung cancer occupies the first place among malignant tumors, accounting for 11.6% of all cases [1]. Among them, non-small cell lung cancer (NSCLC) is the main subcategory of lung cancer [2]. Besides, the high metastasis and recurrence rates of NSCLC result in a 5-year survival rate of approximately 16% [3]. Hence, exploring the potential mechanism of NSCLC is essential for improving NSCLC prognosis.

Unlike linear RNAs, circular RNAs (circRNAs) are endogenous RNAs without 5ʹ to 3ʹ polarity [4]. Extensive studies have identified that highly stable circRNAs participate in different biological processes of various cancers [5]. For example, circBIRC6 silence restrained cell proliferation and metastasis in NSCLC through inhibition of miR-145 [6]. In addition, down-regulation of circFOXM1 hindered NSCLC cell proliferation and cell cycle via mediating the miR-614/FAM83D pathway [7]. Also, circ-SOX4 silencing impeded the malignant evolution of NSCLC through inactivating Wnt/β-catenin signaling [8]. According to high-throughput sequencing (GSE112214), hsa_circ_0060937 derived from cytochrome P450 family 24 subfamily A member 1 (CYP24A1) gene was strikingly elevated in NSCLC tissues. Nonetheless, the potential mechanism of circ_0060937 in NSCLC remains largely unclear.

Extensive investigations have manifested that circRNAs weaken the activity of microRNAs (miRNAs) via serving as competing endogenous RNAs (ceRNAs) [9,10]. In the current research, we speculated that circ_0060937 might interact with miR-195-5p based on prediction software. In triple-negative breast cancer, circAGFG1 contributed to tumor progression via adsorbing miR-195-5p to activate CCNE1 [11]. Furthermore, circRNA_001160 impeded the permeability of the blood-tumor barrier through interacting with miR-195-5p and up-regulating ETV1 [12]. Nevertheless, the interaction between circ_0060937 and miR-195-5p still needs to be confirmed.

Moreover, this research predicted that miR-195-5p might target high-mobility group box 3 (HMGB3). Besides, some studies have illuminated that dysregulated HMGB3 in various cancers is strongly related to tumor progression and chemoresistant [13,14]. For example, depletion of HMGB3 hindered mammosphere formation in breast cancer via binding to HIF1α [15]. Additionally, inhibition of HMGB3 increased cisplatin sensitivity in ovarian cancer cells, providing a promising therapeutic strategy for ovarian cancer chemoresistance [16]. Also, HMGB3 silencing restrained NSCLC cell proliferation and accelerated apoptosis [17].

Hence, we clarified the potential function of circ_0060937 in NSCLC. Further, we also elucidated the potential interaction between circ_0060937 and miR-195-5p/HMGB3 axis in NSCLC progression.

Materials and methods

Clinical samples

This research was endorsed by the Ethics Committee of the First Hospital of Yulin City. Tissue samples composed of NSCLC tissues (n = 65) and matched normal tissues (n = 65) were collected from NSCLC patients undergoing surgery at the First Hospital of Yulin City. Written informed consent was collected from all subjects.

Cell culture and transfection

All cells were cultivated in RPMI-1640 medium (Youkang Biotech, Beijing, China) containing 10% FBS (Youkang Biotech) with 5% CO₂ at 37°C. Among them, four NSCLC cell lines (Calu-3, Calu-6, A549 and H1299) were commercially obtained from American Type Culture Collection (ATCC, Manassas, VA, USA). Additionally, human bronchial epithelial cell line HBE1 was purchased from Procell Life Science&Technology Co., Ltd (Wuhan, China).

Circ_0060937 short hairpin RNA (sh-circ#1, sh-circ#2 and sh-circ#3), miR-195-5p mimics (miR-195-5p), miR-195-5p inhibitor (anti-miR-195-5p), HMGB3 overexpression plasmids (HMGB3) as well as control groups (sh-NC, miR-NC, anti-NC and vector) were synthesized from GenePharma (Shanghai, China). Lipofectamine 3000 (Invitrogen, Carlsbad, CA, USA) was utilized to perform cell transfection when the cell confluence reached ~80%.

Quantitative real-time PCR (qRT-PCR)

Firstly, RNA was extracted using Trizol (Solarbio, Beijing, China). Besides, the nucleus and cytoplasm of A549 and H1299 cells were separated with PARIS kit (Invitrogen). Next, cDNA was synthesized via the specific RT-PCR kit (Takara, Dalian, China). Afterward, the expression level was monitored via SYBR Green PCR Master Mix (Takara). The comparative 2-ΔΔCt method was used to calculate the relative fold-change of target expression, and GAPDH and U6 (for miR-195-5p) were regarded as internal references. All primers were shown in Table 1.

Table 1.

The sequences of primers

Name		Primers for qRT-PCR (5ʹ-3ʹ)
circ_0060937	Forward	GACATTTATCACCAGAATCGGCT
	Reverse	CATCCCCTGCATTCTTCTGGA
CYP24A1	Forward	GGTGGCGAGACTCAGAACG
	Reverse	GTCGTGCTGTTTCTTGAGACC
miR-195-5p	Forward	GGGGTAGCAGCACAGAAAT
	Reverse	TCCAGTGCGTGTCGTGGA
GAPDH	Forward	GCACCGTCAAGGCTGAGAAC
	Reverse	ATGGTGGTGAAGACGCCAGT
U6	Forward	CTCGCTTCGGCAGCACATA
	Reverse	AACGCTTCACGAATTTGCGT
HMGB3	Forward	CCGGGAAGGAAGAAGCAATTC
	Reverse	AGCGGACATCTTGCCCTTTG

CircRNA stability assay

For Actinomycin D treatment, A549 and H1299 cells were exposed to 2 mg/mL Actinomycin D (Solarbio). In addition, RNase R (Solarbio) was incubated with 2 μg RNA for digestion experiment. Then, the abundance of circ_0060937 and CYP24A1 was determined using qRT-PCR.

Cell viability analysis

Following transfection, cells were grown into 96-well plates and cultivated for 0 h, 24 h, 48 h or 72 h. Afterward, 10 μL CCK-8 (Beyotime, Shanghai, China) was injected into each well to incubate cells. After continuing the reaction for 4 h, cell viability was determined via assessing the optical density at 450 nm wavelength with a Microplate Reader (BioTek, Burlington, VT, USA).

Colony formation analysis

Following transfection, cells were injected into 6-well plates. Afterward, the colonies were incubated with formaldehyde and 0.1% crystal violet after cultivation for 14 days. At last, cell proliferation ability was determined via counting the number of colonies under a microscope.

Flow cytometry

Annexin V-FITC/PI Apoptosis Detection kit (Invitrogen) was applied according to the manufacturer’s requirements. Firstly, A549 and H1299 cells were washed with PBS after digestion with trypsin. After reacting with Annexin V-FITC and PI, the apoptotic cells were analyzed using Flow Cytometer (Beckman Coulter, Miami, FL, USA).

For cell cycle assay, A549 and H1299 cells were digested with trypsin and suspended in PBS. Next, the precipitate collected by centrifugation was re-suspended in PBS. After digestion with RNase, the cells were fixed with 1% PI. Subsequently, cell cycle distribution was tested using Flow Cytometer (Beckman Coulter).

Transwell assay

Unlike cell migration analysis, transwell chambers for cell invasion analysis were initially coated Matrigel (Corning, Corning, NY, USA). Briefly, the cells were cultured in FBS-free RPMI-1640 (Youkang Biotech). Meanwhile, RPMI-1640 harboring 10% FBS (Youkang Biotech) was injected into the bottom chamber. Following 24 h of incubation, the transferred cells were reacted with 0.5% crystal violet. Finally, cell migrated and invaded capabilities were examined via counting the number of transferred cells under a microscope at magnification of 100 × .

Measurement of glucose uptake and lactate product

Glucose uptake and lactate product were detected by Glucose and Lactate Assay Kits (Abcam, Cambridge, UK) following the manufacturer’s requirements.

Dual-luciferase reporter assay

The targeting sequence between miR-195-5p and circ_0060937 or HMGB3 3ʹUTR was evaluated through circBANK and starbase online databases. Afterward, the wild-type (WT) vectors of circ_0060937 (circ_0060937-WT) and HMGB3 (HMGB3 3ʹUTR-WT), and the mutant (MUT) plasmids of circ_0060937 (circ_0060937-MUT) and HMGB3 (HMGB3 3ʹUTR-MUT) were built by Genepharma. After that, the luciferase reporter plasmids were firstly mixed with miR-195-5p or miR-NC, followed by treatment with Lipofectamine 3000 (Invitrogen). Finally, Firefly luciferase activity was tested via Dual-Lucy Assay Kit (Solarbio). In addition, Renilla luciferase activity was identified as a reference.

RNA immunoprecipitation (RIP) assay

RIP analysis was implemented using EZ-Magna RIP kit (Millipore, Billerica, MA, USA). After lysing A549 and H1299 cells in RIP lysis buffer, cell lysates were treated with magnetic beads bound with AGO2 antibody or IgG antibody at 4°C. Next, the RNA levels were measured via qRT-PCR assay.

RNA pull-down analysis

Biotinylated circ_0060937 or HMGB3 probe (circ_0060937 probe or Bio-HMGB3-probe) and the control probe (control probe or Bio-NC-probe) were provided from GenePharma. Afterward, A549 and H1299 cells were lysed and reacted with streptavidin-coated magnetic beads (Invitrogen). Subsequently, the contents of six miRNAs and HMGB3 were measured via qRT-PCR analysis.

Western blot assay

Protein was extracted using RIPA lysis buffer (Keygen,Nanjing, China). After detection of protein concentration with BCA protein assay kit (Tiangen, Beijing, China), protein samples were separated by 10% SDS-PAGE and transferred onto PVDF membranes (Millipore, Billerica, MA, USA). Then, the membranes were incubated with anti-HMGB3 (1:2500; ab72544, Abcam) or anti-GAPDH (1:2500; ab9485, Abcam) after blocking with 5% nonfat milk. Subsequently, corresponding secondary antibody (1:20,000; ab205718, Abcam) was used to incubate the membranes. Finally, the protein bands were determined using ECL reagent (Solarbio).

In vivo assay

Firstly, the Animal Ethics Committee of the First Hospital of Yulin City approved in vivo experiments. BALB/c nude mice aged 5 weeks were averagely divided into four groups (n = 5 for each group). Lentivirus with sh-circ#1 or sh-NC was bought from GenePharma and transfected into A549 and H1299 cells. Afterward, stably transfected cells (5 × 10⁶) were subcutaneously injected into the right abdomen of nude mice. Then, tumor volume was measured every 7 days. Until the 35th day after injection, the formed tumor was excised and weighed. Furthermore, the contents of circ_0060937, miR-195-5p and HMGB3 were determined by qRT-PCR and western blot in the excised tumors.

Statistical analysis

Data from three independent repeated tests were tested using Graphpad Prism 7.0 software (GraphPad, San Diego, CA, USA). The differences were evaluated by Student’s t-test or one-way analysis of variance. Statistical significance was considered when P value <0.05.

Results

Circ_0060937 level is increased in NSCLC

First, high-throughput sequencing results presented that circ_0060937 level was overtly elevated in lung cancer tissues compared to adjacent tissues (Figure 1a). Next, qRT-PCR assay illuminated that circ_0060937 level was prominently elevated in NSCLC tissues compared to normal tissues (Figure 1b). Besides, Kaplan-Meier survival analysis unveiled that high level circ_0060937 was related to poor overall survival (Figure 1c). Compared with HBE1 cells, circ_0060937 level was remarkably elevated in NSCLC cells (Calu-3, Calu-6, A549 and H1299 cells) (Figure 1d). As exhibited in Figure 1e, the half-life of circ_0060937 exceeded 24 h, while the half-life of CYP24A1 was less than 16 h after Actinomycin D exposure. Additionally, circ_0060937 was more resistant to RNase R than CYP24A1 (figure 1f). Overall, these data indicated that circ_0060937 might play a carcinogenic effect in NSCLC.

Figure 1.

Circ_0060937 level is increased in NSCLC. (a) Circ_0060937 level in three pairs of lung cancer and normal tissues (GSE112214) was shown. (b) Circ_0060937 level was measured in NSCLC tissues (n = 65) and adjacent normal tissues (n = 65). (c) The association between circ_0060937 level and overall survival was analyzed. (d) Expression of circ_0060937 in HBE1 and NSCLC cells. (e) Following Actinomycin D stimulation, circ_0060937 and CYP24A1 levels were measured at different times. (f) After RNase R stimulation, circ_0060937 and CYP24A1 levels were determined. **P < 0.01, ***P < 0.001

Depletion of circ_0060937 inhibits proliferation and accelerates apoptosis in NSCLC cells

To clarify the function of circ_0060937 in NSCLC, loss-of-function experiments were carried out by transfecting circ_0060937 shRNA into NSCLC cells. First, circ_0060937 was remarkably elevated after introduction of circ_0060937 shRNA, and sh-circ#1 and sh-circ#2 were selected for follow-up experiments (Figure 2a). Additionally, introduction of sh-circ#1 or sh-circ#2 markedly reduced the viability and colony number of A549 and H1299 cells (Figure 2b and Figures 2c). Furthermore, flow cytometry suggested that interference of circ_0060937 significantly facilitated apoptosis and G0/G1 phase arrest in NSCLC cells (Figure 2d and Figure 2e). These data evidenced that circ_0060937 deletion hindered proliferation and accelerated apoptosis in NSCLC cells.

Figure 2.

Depletion of circ_0060937 inhibits proliferation and accelerates apoptosis in NSCLC cells. (a) Circ_0060937 level was tested in NSCLC cells following circ_0060937 shRNA transfection. (b and c) Cell proliferation was evaluated by detecting cell viability and colony number. (d and e) Cell apoptosis and cycle distribution were detected via flow cytometry. *P < 0.05, **P < 0.01

Circ_0060937 silencing hinders the migration, invasion and glycolysis of NSCLC cells

Besides, transwell analysis was applied to investigate the effect of circ_0060937 on NSCLC cell metastasis. As presented in Figure 3a-Figure 3b, transfection with sh-circ#1 or sh-circ#2 prominently decreased the migrated and invaded cells in comparison with the control group. Furthermore, glycolysis promotes the energy supply of tumors by increasing the oxidation of glucose. As displayed in Figure 3c, depletion of circ_0060937 led to a conspicuous reduction in glucose uptake and lactate product. Collectively, these results indicated that circ_0060937 silence restrained the metastasis and glycolysis of NSCLC cells.

Figure 3.

Circ_0060937 silencing hinders the metastasis and glycolysis of NSCLC cells. sh-circ#1, sh-circ#2 or sh-NC was introduced into A549 and H1299 cells. (a and b) Cell migration and invasion capabilities were tested via transwell analysis. (c) After transfection for 48 h, glycolysis was determined by assessing glucose uptake and lactate product. **P < 0.01

Circ_0060937 directly interacts with miR-195-5p

First of all, the results exhibited that circ_0060937 was principally distributed in the cytoplasm of NSCLC cells (Figure 4a). RIP assay revealed that circ_0060937 bound to AGO2 protein (Figure 4b). Next, circBANK (http://www.circbank.cn/index.html) and starbase (http://starbase.sysu.edu.cn/) online databases predicted potential targets of circ_0060937 (Figure 4c). Pull-down analysis suggested that miR-195-5p was strikingly enriched by the circ_0060937 probe, but other miRNAs (miR-497-5p, miR-424-5p, miR-16-5p, miR-15a-5p and miR-15b-5p) were not pulled down (Figure 4d and Figure 4e). As shown in figure 4f, online databases displayed the possible binding sequence between miR-195-5p and circ_0060937. Further, dual-luciferase reporter assay confirmed that miR-195-5p mimics overtly reduced the luciferase activity of circ_0060937-WT (Figure 4g and Figure 4h). Moreover, miR-195-5p level was strikingly declined in NSCLC tissues relative to the control group (Figure 4i). Spearman’s correlation coefficient depicted that circ_0060937 was negatively correlated with miR-195-5p in NSCLC tissues (Figure 4j). Also, miR-195-5p level was remarkably declined in NSCLC cells compared to HBE1 cells (Figure 4k). Overall, these data illuminated that circ_0060937 sponged miR-195-5p.

Figure 4.

Circ_0060937 directly interacts with miR-195-5p. (a) qRT-PCR analysis determined the distribution of circ_0060937 in NSCLC cells. (b) The relationship between circ_0060937 and AGO2 was analyzed using RIP assay. (c) Venn diagram displayed some miRNAs with binding sites to circ_0060937 in circBANK and starbase databases. (d and e) RNA pull-down analysis was utilized to confirm the possible miRNA targets for circ_0060937. (f) The complementary sequence of miR-195-5p in circ_0060937 was exhibited. (g and h) Dual-luciferase reporter assay was utilized to validate the binding relationship. (i) MiR-195-5p level was examined in NSCLC and normal tissues. (j) Spearman’s correlation analysis was applied to verify the association between circ_0060937 and miR-195-5p. (k) MiR-195-5p expression was tested in HBE1 cells and NSCLC cells. **P < 0.01, ***P < 0.001

Knockdown of miR-195-5p alleviates the inhibition of circ_0060937 silence on NSCLC cell progression

To investigate whether circ_0060937 affected NSCLC development via mediating miR-195-5p, a range of rescue experiments were carried out in A549 and H1299 cells introduced with sh-circ#1 or/and anti-miR-195-5p. First, the suppression efficiency of anti-miR-195-5p was validated using qRT-PCR analysis (Figure 5a). CCK-8 and colony formation assays showed that circ_0060937 silence inhibited proliferation of A549 and H1299 cells, which was abolished after introduction of anti-miR-195-5p (Figure 5b and Figure 5c). Additionally, inhibition of circ_0060937 facilitated apoptosis and cell cycle arrest in A549 and H1299 cells, whereas down-regulation of miR-195-5p alleviated these changes (Figure 5d and Figure 5e). Furthermore, silencing of circ_0060937 restrained cell migration, invasion and glycolysis, while these effects were abated by inhibiting miR-195-5p (Figure 5f-Figure 5i).These data evidenced that circ_0060937 depletion blocked NSCLC progression by modulating miR-195-5p.

Figure 5.

Knockdown of miR-195-5p alleviates the inhibition of circ_0060937 silence on NSCLC cell progression. (a) MiR-195-5p level was detected in A549 and H1299 cells introduced with anti-NC or anti-miR-195-5p. (b-i) A549 and H1299 cells were introduced with sh-circ#1 or/and anti-miR-195-5p. (b and c) Cell proliferation ability was assessed via CCK-8 and colony formation assays. Cell apoptosis and cycle (d and e), cell migration and invasion (f and g), glucose uptake and lactate product (h and i) were determined using the corresponding methods. *P < 0.05, **P < 0.01

Circ_0060937 regulates HMGB3 expression by sponging miR-195-5p

To further investigate the underlying mechanism of circ_0060937, starBase online database displayed that miR-195-5p possessed a putative binding site in HMGB3 3ʹUTR (Figure 6a). Dual-luciferase reporter analysis exhibited that miR-195-5p up-regulation remarkably dropped the luciferase activity of HMGB3 3ʹUTR-WT (Figure 6b). Then, RNA pull-down assay and RIP assay were performed to confirm whether HMGB3 bound to miR-195-5p. As shown in Figure 6c, HMGB3 and miR-195-5p were enriched by Bio-HMGB3-probe, but not by Bio-NC-probe. RIP analysis displayed that HMGB3 and miR-195-5p were conspicuously enriched in anti-AGO group rather than anti-lgG group (Figure 6d). Moreover, RIP assay also indicated that knockdown of circ_0060937 decreased the endogenous enrichment of HMGB3 and miR-195-5p in anti-AGO group (Figure 6e). Besides that, miR-195-5p up-regulation markedly reduced HMGB3 level, and miR-195-5p inhibition strikingly elevated HMGB3 level (figure 6f and Figure 6g). Additionally, knockdown of miR-195-5p reversed the decrease in HMGB3 level caused by circ_0060937 silencing (Figure 6h-Figure 6j). Compared with the normal tissues, HMGB3 mRNA level was remarkably increased in NSCLC tissues (Figure 6k). As illustrated in Figure 6l-Figure 6m, HMGB3 was negatively correlated with miR-195-5p and positively correlated with circ_0060937 in NSCLC tissues (Figure 6l and Figure 6m). Altogether, these data indicated that circ_0060937 elevated HMGB3 via absorbing miR-195-5p.

Figure 6.

Circ_0060937 regulates HMGB3 expression by sponging miR-195-5p. (a) The predicted binding sequence for miR-195-5p and HMGB3 3ʹUTR was shown. (b-d) The binding relationship was demonstrated via dual-luciferase reporter, RNA pull-down and RIP assays. (e) The enrichment of miR-195-5p and HMGB3 in A549 and H1299 cells transfected with sh-NC or sh-circ#1 using anti-Ago2. (f and g) HMGB3 mRNA and protein levels were measured in A549 and H1299 cells after miR-195-5p overexpression or miR-195-5p knockdown. (h-j) After introduction with sh-circ#1 or/and anti-miR-195-5p, HMGB3 expression was examined. (k) HMGB3 mRNA level was measured in NSCLC and normal tissues. (l and m) The association among circ_0060937, miR-195-5p and HMGB3 were evaluated via Spearman’s correlation analysis. *P < 0.05, **P < 0.01, ***P < 0.001

HMGB3 overexpression attenuates the repressive effect of circ_0060937 knockdown on NSCLC cell progression

To explore the interaction between HMGB3 and circ_0060937, A549 and H1299 cells were transduced with sh-circ#1 or/and HMGB3. First, HMGB3 level was strikingly elevated following HMGB3 transfection (Figure 7a). As displayed in Figure 7b-Figure 7c, up-regulation of HMGB3 partially attenuated the repressive effect of circ_0060937 interference on NSCLC cell proliferation. Furthermore, introduction of sh-circ#1+ HMGB3 partially abated the promoting effect of circ_0060937 knockdown on apoptosis and cell cycle arrest in A549 and H1299 cells (Figure 7d and Figure 7e). Besides, HMGB3 overexpression partially alleviated the suppression of circ_0060937 depletion on NSCLC cell migration, invasion and glycolysis (figure 7f-Figure 7i). Overall, these data evidenced that knockdown of circ_0060937 restrained NSCLC development via regulating HMGB3.

Figure 7.

HMGB3 overexpression attenuates the repressive effect of circ_0060937 knockdown on NSCLC cell progression. (a) After introduction of vector or HMGB3, HMGB3 protein level was measured using western blot. Cell proliferation ability (b and c), apoptosis (d), cell cycle (e), cell migrated and invaded capabilities (f and g), glucose uptake and lactate product (h and i) were tested via the corresponding methods in A549 and H1299 cells introduced with sh-circ#1 or/and HMGB3 via corresponding methods. *P < 0.05, **P < 0.01

Depletion of circ_0060937 inhibits tumor growth in vivo

We established a xenograft mouse model to analyze the role of circ_0060937 in tumorigenesis. As illustrated in Figure 8a and Figure 8b, tumor volume was remarkably reduced in the sh-circ#1 group compared with the sh-NC group. Similarly, introduction of sh-circ#1 significantly decreased tumor weight (Figure 8c and Figure 8d). Moreover, circ_0060937 and HMGB3 levels were strikingly reduced, while miR-195-5p level was overtly increased in the sh-circ#1 group relative to the sh-NC group (Figure 8e-Figure 8g). These data illuminated that circ_0060937 silence hindered tumorigenesis in vivo.

Figure 8.

Depletion of circ_0060937 inhibits tumorigenesis in vivo. (a and b) Tumor volume was examined every 7 days. (c and d) Until the 35th day after injection, the formed tumor was excised and weighed. (e-g) The contents of circ_0060937, miR-195-5p and HMGB3 in xenograft tumors were tested via qRT-PCR and western blot. **P < 0.01

Discussion

Glycolysis is a pivotal hallmark of tumor energy metabolism, and lactate is a crucial medium for metabolic communication [18]. Some key factors in glycolysis-related pathways are closely related to tumor malignancy and drug resistance [19]. Compelling evidence has manifested that circRNA exerts a vital effect on regulating cell metabolism via its unique structure [20]. For instance, circAKT3 facilitated glycolysis to reduce cell sensitivity to cisplatin in lung carcinoma via sponging miR-516b-5p to up-regulate STAT3 [21]. Besides, previous literatures have demonstrated that many circRNAs occupy an important position in NSCLC progression [5]. For example, circ_0087862 contributed to NSCLC tumorigenesis via emerging as a decoy for miR-1253 to elevate RAB3D expression [22]. Wu et al. unveiled that circ-ACACA expedited NSCLC development and glycolysis by binding to miR-1183 and modulating PI3K/PKB pathway [23]. In this research, circ_0060937 level was overtly increased in NSCLC, and circ_0060937 depletion impeded NSCLC progression and glycolysis.

Accumulating investigations have corroborated that circRNAs serve as miRNA sponges to affect tumor progression. In gastric cancer, circRNA_100876 accelerated tumor growth via down-regulating miR-136 to activate MIEN1 [24]. In nasopharyngeal carcinoma, circ-ZNF609 depletion blocked tumor malignancy by serving as a ceRNA for miR-188 to decline ELF2 [25]. In our research, we verified that miR-195-5p directly interacted with circ_0060937. Meanwhile, miR-195-5p was prominently down-regulated in NSCLC. Previous studies validated that miR-195-5p exerted an anticancer effect. Pan et al. indicated that overexpression of miR-195-5p impeded the metastasis of cervical carcinoma through down-regulation of ARL2 [26]. Wang et al. disclosed that augmentation of miR-195-5p blocked oral squamous cell carcinoma development via inhibiting TRIM14 [27]. Furthermore, miR-195-5p targeted CEP55 to restrain NSCLC progression via regulating cell proliferation and apoptosis [28]. Herein, circ_0060937 knockdown impaired NSCLC development and glycolysis by absorbing miR-195-5p, which has not been clarified in previous studies.

Recently, mounting studies have evidenced that miRNAs regulate a range of biological functions through modulating target genes [29]. According to bioinformatics analysis, we speculated that miR-195-5p directly interacted with HMGB3. In bladder cancer, suppression of miR-532-5p aggravated tumor malignancy via interacting with HMGB3 through Wnt/β-catenin pathway [30]. Additionally, miR-758 triggered tumor growth and metastasis in cervical carcinoma via binding to HMGB3 through mediating Wnt/β-catenin signaling [31]. In prostate carcinoma, miR-205-5p negatively regulated HMGB3 to implicating in tumor progression [32]. Additionally, up-regulation of HMGB3 accelerated cell growth and metastasis in NSCLC by binding with miR-513b [33]. In this research, our findings unveiled that miR-195-5p negatively regulated HMGB3. More importantly, circ_0060937 silencing hindered NSCLC progression and glycolysis by modulating HMGB3, and circ_0060937 sponged miR-195-5p to increase HMGB3 expression.

In conclusion, circ_0060937 expedited NSCLC progression via absorbing miR-195-5p and up-regulating HMGB3. This research hinted that circ_0060937/miR-195-5p/HMGB3 axis might be a promising strategy for NSCLC therapy. Nevertheless, our research has some limitations, the expression and function of circ_0060937 in healthy cell lines should be examined to ensure its safety and efficacy before clinical application.

Funding Statement

This work was supported by Natural Science Basic Research Program of Shaanxi (Program No.2018JM7068) and Science and Technology Research and Development Project of Yulin City (Project Number [2019] No. 185-35).

Disclosure statement

The authors declare that they have no financial conflicts of interest.