Introduction
Gastric cancer (GC) is the fourth most usual cancer that endangers human health worldwide [
1,
2]. Despite various improvements have been made in diagnosis and treatment in the past few decades, the prognosis of GC is still very poor due to the metastasis and recurrence of tumors [
3,
4]. Hence, it is imperative to discover new targets for the diagnosis and therapy of GC.
Circular RNAs (circRNAs) are a class of non-coding RNAs (ncRNAs) formed by back-splicing and implicated in human diseases, especially cancers [
5‐
7]. Moreover, emerging evidence has displayed that circRNAs can act as microRNA (miRNA) sponges to influence gene expression, thereby regulating the biological processes in tumor cells [
8,
9]. For example, circ_0067934 facilitated cell invasion and growth and repressed apoptosis in thyroid cancer by elevating CXCR1 through sponging miR-1304 [
10]. Circ-ADAM9 aggravated the growth and motility of pancreatic cancer via miR-217/PRSS3 axis [
11]. The involvement of circRNAs in GC has also been widely explored [
12‐
14]. CircRNA ring finger protein 111 (circ-RNF111, hsa_circ_0001982) was identified as an oncogene in breast cancer [
15], colorectal cancer [
16] as well as GC [
17]. Even so, the functions and regulatory mechanisms of circ-RNF111 remain largely unknown.
MiRNAs are short ncRNAs and served as tumor promoters or suppressors in GC [
18]. Yang et al. suggested that miR-876-3p enhanced the apoptosis and curbed the proliferation of pancreatic adenocarcinoma by targeting JAG2 [
19]. Tang et al. declared that miR-876-3p directly targeted KIF20A to slow the carcinogenesis of glioma [
20]. Moreover, the downregulation of miR-876-3p was related to the worse outcome of GC patients, and enhanced chemoresistance of GC by targeting TMED3 [
21]. Nevertheless, the exact functions of miR-876-3p in GC remain unclarified.
Krueppel-like factor 12 (KLF12) is a member of the KLFs’ family and plays essential roles in various human cancers, including GC [
22,
23]. In the study, bioinformatics analysis presented that miR-876-3p included the binding sequences of circ-RNF111 and KLF12, indicating the potential relationships of circ-RNF111, miR-876-3p, and KLF12.
Here, the expression pattern of circ-RNF111 in GC was determined. Moreover, the functions and relationships of circ-RNF111, miR-876-3p, and KLF12 in GC development were investigated.
Materials and methods
Tissues acquisition
Thirty-one GC patients at the First People’s Hospital of Xiaoshan were recruited in our research. After the study was permitted by the Ethics Committee of the First People’s Hospital of Xiaoshan and written informed consents were offered by the participants, the tumor tissues and adjacent non-tumor tissues were acquired and stored at – 80 °C before use.
Cell culture
GES-1 cells and AGS cells were obtained from Procell (Wuhan, China). GC cells (SNU-638) were acquired from Shanghai GuanDao Biological Engineering Co., Ltd. (Shanghai, China). The cells were cultivated at 37°C in DMEM (Sigma-Aldrich, St. Louis, MO, USA) added with 10% FBS (Sigma-Aldrich) and 1% penicillin-streptomycin (Sigma-Aldrich) in a humidified incubator.
Quantitative real-time polymerase chain reaction (qRT-PCR)
After being extracted via RNAiso Plus (Takara, Dalian, China), the RNA was reversely transcribed into cDNAs with PrimeScript™ RT reagent Kit (Takara) or TaqMan miRNA assays (Applied Biosystems, Foster City, CA, USA). Then, qRT-PCR was conducted through the usage of SYBR Premix DimerEraser (Takara) and related primers (Sangon, Shanghai, China). The expression was estimated with the 2−ΔΔCt way with normalization to GAPDH or U6. The primers included: circ-RNF111: (F: 5′-ACAATCCAGCTGTTCCCTCA-3′ and R: 5′-GGCTCTGGATGCAAAAGGAT-3′); miR-876-3p: (F: 5′-CTGTGGTGGTTTACAAAGTAATT-3′ and R: 5′-GTGCAGGGTCCGAGGT-3’); KLF12: (F: 5′-TGGCAAAGCACAAATGGAC-3′ and R: 5′-CTAAATGGTGAAATTGAACAAGG-3′); GAPDH: (F: 5′-GGAGTCCACTGGCGTCTTCA-3′ and R: 5′-GGTTCACACCCATGACGAAC-3′); U6: (F: 5′-ATTGGAACGATACAGAGAAGATT-3′ and R: 5′-GGAACGCTTCACGAATTTG-3′). To analyze the feature of circ-RNF111, total RNA was exposed to RNase R (3 U/μg; Epicentre, Madison, WI, USA) for 15 min at 37 °C and then circ-RNF111 and GAPDH levels were detected.
Cell transfection
Circ-RNF111 siRNAs (si-circ-RNF111#1 and si-circ-RNF111#2), circ-RNF111 overexpression plasmid (circ-RNF111), miR-876-3p mimics, miR-876-3p inhibitors (anti-miR-876-3p), KLF12 overexpression plasmid (KLF12), circ-RNF111 shRNA (sh-circ-RNF111), and related controls (si-NC, Vector, miR-NC, anti-miR-NC, pcDNA, and sh-NC) were designed by GenePharma (Shanghai, China). GC cells were introduced with the compositions utilizing Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA).
Cell counting kit-8 (CCK-8) assay
Firstly, 5 × 103 cells were seeded into each well of 96-well plates. After 48 h of incubation, 10 μL CCK-8 (Sigma-Aldrich) was supplemented into the well and kept for an additional 2 h. At last, the absorption at 450 nm was measured.
Following indicated transfection, GC cells were seeded into 6-well plates for 14 days. When the colonies were visible, the culture was terminated. Next, the colonies were dyed with crystal violet (Sangon) and quantified utilizing a microscope (Olympus, Tokyo, Japan).
Flow cytometry analysis
The apoptosis and cell cycle process were examined with Annexin V-FITC/PI Apoptosis Kit (Beyotime, Shanghai, China) based on the manufacturers’ instructions. To examine cell apoptosis, GC cells with various transfections were harvested, washed with PBS (Sangon), and then resuspended in binding buffer. Then, Annexin V-FITC and PI were adopted to dye the cells. For cell cycle process, after the transfected cells were washed, they were fixed with 70% ethanol and then kept with RNase (Solarbio, Beijing, China) in PBS (Sangon) for 1 h. Thereafter, the cells were interacted with PI. The apoptotic rate and cell cycle were assessed with a FACS flow cytometry (BD Biosciences, Franklin Lakes, NJ, USA).
Wound-healing assay
To determine the migration of GC cells, the transfected GC cells were added into 6-well plates and grown until 100% confluence. Then, the sterile pipette tip was utilized to make a scratch in the well. At 0 h and 24 h, the crosses were recorded.
Transwell assay
The invasion of GC cells was evaluated with the transwell insert chambers (BD Biosciences) pre-covered Matrigel (BD Biosciences). In short, the transfected AGS and SNU-638 cells in DMEM (Sigma-Aldrich) with serum-free were added into the upper chamber. The bottom chamber was filled with DMEM (Sigma-Aldrich) including 10% FBS (Sigma-Aldrich). Twenty-four hours later, the invaded cells were dyed with crystal violet (Sangon) and quantified utilizing a microscope (Olympus; × 100 magnification).
Measurement of glycolysis level
The lactate assay kit (Sigma-Aldrich), glucose assay kit (Sigma-Aldrich), and ATP assay kit (Sigma-Aldrich) were employed to test glucose uptake, lactate production, and ATP production levels according to the manufacturers’ protocols.
Western blot assay
The protein was contained by lysing tissues and cells into RIPA buffer (Beyotime), electrophoresed on SDS-PAGE (Solarbio, Beijing, China) and then blotted onto PVDF membranes (Millipore, Billerica, MA, USA). After blockage in 5% skim milk for 1 h, the proteins were interacted overnight with primary antibodies and secondary antibody (bs-0295M-HRP; Bioss, Beijing, China) for 1.5 h. The bands were visualized with ECL reagent (Beyotime) and band intensity was examined with ImageJ (NIH, Bethesda, MD, USA). The primary antibodies including β-actin (bs-0061R; Bioss), hexokinase2 (HK-2; bs-3993R; Bioss), CyclinD1 (bs-0623R; Bioss), MMP9 (bs-4593R; Bioss), or KLF12 (bs-16783R; Bioss)
Dual-luciferase reporter assay
The fragments of wild-type (WT) circ-RNF111, or KLF12 3′UTR including miR-876-3p binding sites or mutant (MUT) circ-RNF111 or KLF12 3′UTR lacking miR-876-3p binding sequences were introduced into pmirGLO (Promega, Fitchburg, WI, USA) to generate WT-circ-RNF111, MUT-circ-RNF111, WT-KLF12 3′UTR, and MUT-KLF12 3′UTR, respectively. Next, the generated vectors and miR-NC/miR-876-3p were administrated into GC cells. The luciferase activity was measured utilizing Dual-Luciferase Reporter Assay System (Promega).
RNA immunoprecipitation (RIP) assay
In brief, after AGS and SNU-638 cells were lysed in RIP buffer, the cell lysates were maintained with protein A/G sepharose beads conjugated with antibody IgG or Ago2. Next, total RNA in immunoprecipitates was subjected to qRT-PCR for the abundance of circ-RNF111, miR-876-3p, and KLF12.
Murine xenograft model
The BALB/c nude mice from Beijing Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China) were assigned into 2 groups (n = 5/group). Sh-NC or sh-circ-RNF111 transfected AGS cells (2 × 105) were suspended in 0.2 mL PBS (Sangon) and then subcutaneously introduced into the flank of the mice. After 7 days, tumor size was examined every 5 days and estimated via the formula: (LengthWidth2) × 0.5. At 32 days, the mice were sacrificed and xenograft tumors were weighted. The in vivo study obtained permission from the Ethics Committee of Animal Research of the First People’s Hospital of Xiaoshan.
Immunohistochemistry (IHC) assay
The expression of KLF12 and ki67 in the xenograft tumor tissues was examined by IHC assay, as previously described [
24]. The antibodies against KLF12 (bs-16783R) and ki67 (bs-23103R) were provided by Bioss.
Statistical analysis
The results from three independent experiments were analyzed using GraphPad Prism 7 and exhibited as mean ± SD. The relationship between miR-876-3p level and circ-RNF111 level or KLF12 level in GC tissues was evaluated by Spearman’s correlation coefficient. Student’s t test or one-way analysis of variance was utilized for different analysis. P < 0.05 was thought to be significant.
Discussion
Recently, circRNAs have attracted researchers’ attention for their potential in cancer biology [
25]. Substantial evidence implied that circRNAs are engaged in enhancing or repressing GC development. For instance, circ_001653 promoted GC malignancy by upregulating NR6A1 via decoying miR-377 [
26]. Circ_0027599 directly targeted miR-101-3p.1, leading to the suppression in GC cell growth and metastasis [
14]. Herein, we clarified the functions of circ-RNF111 in GC development. As a result, the upregulation of circ-RNF111 triggered the malignant phenotypes of GC cells. Furthermore, we discovered a novel pathway of circ-RNF111/miR-876-3p/KLF12 in regulating GC development.
Tang et al. manifested that circ_0001982 was overexpressed in breast cancer, and promoted tumor cell growth and invasion through adsorbing miR-143 [
15]. Deng et al. demonstrated the oncogenic role of circ_0001982 in colorectal cancer development through sponging miR-144 [
16]. Moreover, Wang et al. uncovered that circ-RNF111 sponged miR-27b-3p to aggravate GC cell growth and metastasis and repress apoptosis [
17]. Corresponding to the previous studies, we also demonstrated the promotional effect circ-RNF111 in GC. In the present research, circ-RNF111 was elevated in GC. Functionally, circ-RNF111 interference curbed GC cell viability, colony formation, motility, triggered apoptosis and blocked cell cycle process. Besides, tumor cells prefer to obtain energy to meet the need for their rapid growth rather than oxidative phosphorylation and the suppression of glycolysis plays a vital role to hamper tumor progression [
27,
28]. Therefore, we explored glycolysis level in GC cells and found that circ-RNF111 silencing decreased glucose uptake, lactate production, ATP synthesis and HK-2 levels, thereby suppressing glycolysis. In addition, to further explore the function of circ-RNF111, the murine xenograft model was established. It was demonstrated that circ-RNF111 silencing restrained tumor formation in vivo. All these findings demonstrated the promotional effect of circ-RNF111 on GC malignancy.
Mechanistically, circ-RNF111 was identified to promote KLF12 expression by serving as the sponge for miR-876-3p in GC cells. MiR-876-3p has been demonstrated to be targeted by circ_0088494 and circTP53 [
29,
30], but the association between circ-RNF111 and miR-876-3p was the first to be elucidated. Herein, miR-876-3p was reduced in GC patients MiR-876-3p overexpression repressed GC cell growth, metastasis, and glycolysis and accelerated apoptosis. Moreover, miR-876-3p inhibitor reversed circ-RNF111 knockdown-mediated suppressive roles in GC cell malignant behaviors, indicating that circ-RNF111 influenced GC progression by decoying miR-876-3p. The oncogenic function of KLF12 has been elucidated in several cancers, such as colorectal cancer [
31], pancreatic cancer [
32], osteosarcoma [
33], and nasopharyngeal carcinoma [
34]. Moreover, KLF12 could be targeted by miR-618 [
35], miR-200a-3p [
22], miR-376b-3p [
36], and miR-137 [
37]. In this work, we discovered that miR-876-3p directly interacted with KLF12 to negatively modulate KLF12 expression in GC cells. The impacts of miR-876-3p overexpression in GC cell malignant phenotypes were abrogated by increasing KLF12, suggesting miR-876-3p could target KLF12 to alter GC progression.
In conclusion, circ-RNF111 was abnormally increased in GC. Moreover, circ-RNF111/miR-876-3p/KLF12 axis could deteriorate the progression of GC by triggering tumor cell growth, metastasis, and glycolysis and curbing apoptosis. Our results might offer key clues to develop an effective treatment method for GC.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.