Introduction
Gastric cancer (GC) is the fifth most common cancer and ranks as the third major cause of cancer-related death globally [
1,
2]. Due to the lack of specific symptoms and diagnostic markers in early stage GC, GC patients are always diagnosed at an advanced stage [
3,
4]. Although treatments such as molecular-targeted therapy have emerged, surgical resection and adjuvant chemotherapy still serve as the main treatments for GC [
5]. Chemotherapy is a first-line regimen with a poor curative effect due to chemoresistance, which develops by genetic and epigenetic modifications, signaling pathway alterations or cell metabolism disorders [
6]. Therefore, further investigation of the mechanism of carcinogenesis and chemoresistance in GC is of great concern for early diagnosis and for improving the prognosis and survival of patients with advanced GC.
Circular RNA (circRNA) is a newly discovered type of noncoding RNA (ncRNA) characterized by a special single-stranded closed loop lacking both 5′-3′ polarity and a polyadenylated tail [
7]. circRNA expression was found to be specific in diverse cell lines and tissue types, implying that circRNA has multiple functional roles in biological and pathological processes [
8,
9]. An increasing number of studies have shown that circRNAs are differentially expressed in various cancers, such as breast cancer, oral squamous cell carcinoma and gastric cancer [
10‐
12]. Studies have shown that circRNAs participate in various biological processes, such as proliferation, migration, apoptosis and the cell cycle, and thus participate in tumorigenesis and chemoresistance. Additionally, autophagy, which participates in carcinogenesis and chemoresistance, could be modulated by circRNA [
13‐
15].
In this study, we found that circCUL2 was downregulated in GC tissues and cell lines. circCUL2 (circbase ID: hsa_circ_0000234), derived from back-splicing of the CUL2 mRNA (from exon 2 to exon 4), is located on chromosome 10:35,349,801-35,360,267 and is 339 nucleotides (nt) in length. It suppressed the proliferation, migration and invasion of GC cells. Furthermore, circCUL2 was downregulated in cisplatin-resistant GC cell lines and modulated cisplatin sensitivity. circCUL2 may be involved in tumorigenesis and chemoresistance by competitively binding to miR-142-3p and by modulating ROCK2 expression and autophagy activation.
Methods
Clinical specimens
Microarray data on 5 paired GC patients (GSE100170) [
16] were downloaded from the GEO database and analyzed by R software. One hundred pairs of GC tissues and paired normal tissues were obtained from patients who underwent surgery at Jiangsu Province Hospital between 2011 and 2017. None of the patients received chemotherapy or radiotherapy before surgery. All the sample diagnoses were confirmed through pathological analysis; then, the samples were frozen in liquid nitrogen for 15 min and stored at − 80 °C until use. Blood samples were collected from 48 GC patients and 48 healthy controls (with no diagnosis of cancer). Informed consent was obtained from patients involved in the study. Tumors were staged according to the tumor-node-metastasis (TNM) staging system of the International Union Against Cancer (v.8, 2016). The protocols of this study were approved by the Ethics Committee of First Affiliated Hospital of Nanjing Medical University.
Cell culture and transfection
Human GC cell lines (AGS, SGC-7901, MKN-45, and BGC-823) and the normal human gastric epithelial cell line GES-1 were purchased from the American Type Culture Collection (ATCC, USA). All these cells were cultured in RPMI 1640 medium (GIBCO, Brazil) supplemented with 10% fetal bovine serum (FBS; GIBCO, Brazil) and incubated at 37 °C in a humidified atmosphere with 5% CO
2. AGS and SGC-7901 cells were transfected with Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA). After transfection for 24 h, cells were harvested for further investigation. The sequences used are listed in Additional file
1: Table S1. We synthesized human circCUL2 cDNA and subsequently cloned it into the pcDNA3.1 vector (Thermo Fisher, USA) to construct the circCUL2 overexpression plasmid.
RNase R digestion and actinomycin D assay
A total of 3 μg of RNA was incubated with 20 U/μL RNase R (Epicentre Biotechnologies) for 15 min at 37 °C to confirm the circRNA characteristics. AGS and SGC-7901 cells were treated with actinomycin D for 0 h, 4 h, 8 h, 12 h and 24 h before RNA extraction. circCUL2 and CUL2 were detected in these two cell types.
Sanger sequencing
The amplification products of circRNAs were inserted into a T-vector for Sanger sequencing by Tsingke (Nanjing, China). The primer (Invitrogen, Shanghai, China) was designed to confirm the back-splice junction of circCUL2.
RNA extraction and quantitative real-time PCR (qRT-PCR)
Total RNA was isolated from clinical specimens or cell lines with TRIzol reagent (Invitrogen, Carlsbad, CA, USA). RNA (500 ng) was reverse transcribed into cDNA following the protocol of the PrimeScript RT Reagent Kit (TaKaRa Bio, Inc., China). Real-time quantitative PCR was used to detect the expression levels of circCUL2, miR-142-3p, and ROCK2. The circRNA and mRNA levels were normalized to that of GAPDH. The miRNA level was normalized to that of U6. The PCR primer sequences were synthesized by Tsingke (Nanjing, China) and are listed in Additional file
1: Table S2. RNA expression fold changes were determined with the 2
−ΔCt method.
Isolation of nuclear and cytoplasmic fractions
Cytoplasmic and nuclear fractions were prepared using NE-PER Nuclear and Cytoplasmic Extraction Reagents (Thermo Scientific, USA) according to the manufacturer’s protocol. GC cells were lysed on ice for 10 min in Lysis Buffer J supplemented with protease inhibitors. After centrifugation at 14,000×g for 3 min, the resulting supernatant and sediment were collected as the cytoplasmic and nuclear fractions, respectively. RNA was extracted from each fraction by Buffer SK and washed with a wash solution. qRT-PCR was used to test the RNA expression of several RNAs.
Fluorescence in situ hybridization (FISH)
Specific probes against circCUL2 and miR-142-3p were used for fluorescence in situ hybridization. The FISH procedures were conducted following the manufacturer’s instructions (GenePharma, Shanghai, China). Briefly, after fixation with 4% paraformaldehyde (PFA) for 15 min at room temperature, cells were washed twice with PBS, and were then mixed with 70%, 95 and 100% ethanol overnight at 4 °C. Hybridization of cells was carried out at 37 °C overnight in a dark moist chamber. After washing three times in saline-sodium citrate (SSC) buffer for 5 min and incubating in blocking buffer (1% BSA and 3% normal goat serum in PBS) for 1 h, cells were incubated with an HRP-conjugated anti-biotin antibody at 4 °C overnight. Finally, the cells were photographed using a fluorescence microscope (Olympus BX53, Olympus America, Inc., Center Valley, PA, USA).
Protein extraction and western blot analysis
Radioimmunoprecipitation assay (RIPA) buffer was used for protein extraction. The supernatants from cell lysates were run on 10% acrylamide gels by SDS-PAGE and then transferred to a polyvinylidene difluoride membrane (Millipore). Antibodies against ROCK2 (1:1000, #ab71598, Abcam), LC3 (1:1000, #12741, Cell Signaling Technology), p62 (1:1000, #ab56416, Abcam), and Beclin1 (1:1000, #ab210498, Abcam) and an HRP-conjugated secondary antibody (1:2000) were employed for western blotting. A chemiluminescence western blotting detection system (Bio-Rad) was used for protein detection.
Immunohistochemistry (IHC)
Tumor tissues were first fixed in 4% paraformaldehyde and embedded in paraffin, and 5-μm-thick sections were cut. Sections were blocked with 10% goat serum and incubated with an anti-ROCK2 (1:1000, #ab71598, Abcam) antibody at 4 °C overnight. Finally, images were acquired for analysis.
Cell proliferation assay
Cell counting kit-8 (CCK-8, Dojindo, Osaka, Japan) was used to measure the proliferative capacity of GC cell lines. AGS and SGC-7901 cells were seeded in 96-well plates. Each well was treated with 10 μL of CCK-8 reagent, followed by incubation at 37 °C for another 1 h before the detection of absorbance at 450 nm by use of a microplate reader (Bio-Rad, CA, USA). Cell proliferation was observed at different times (6, 24, 48, 72 and 96 h). All experiments were performed in triplicate.
5-Ethynyl-2′-deoxyuridine (EdU) incorporation assay
A Cell-Light EdU DNA Cell Proliferation Kit (RiboBio, Guangzhou, China) was used to perform the EdU assay. After incubation with 50 mM EdU for 2 h, the AGS and SGC-7901 cells were fixed in 4% paraformaldehyde and stained with Apollo Dye Solution, and then Hoechst 33342 was used to identify the nuclei. Then, the proliferation-positive cells were photographed and counted under an Olympus FSX100 microscope (Olympus, Tokyo, Japan).
GC cell lines (AGS and SGC-7901) resuspended to 1× 103 cells/mL were seeded in 6-well plates. After incubation at 37 °C for 14 days, cells were stained with 0.1% crystal violet and 20% methanol, and then the cell colonies were counted.
Migration and invasion assay
Transwell assays were used for migration and invasion assays. For invasion assays, the lower chambers were precoated with 100 μL of Matrigel (BD Bioscience, San Jose, CA, USA) for 30 min before the addition of medium to the chambers. The transfected GC cells (1 × 106 cells/mL) were resuspended in RPMI 1640 medium. The upper chamber contained 100 μL of cell suspension medium, and 600 μL of complete medium was added to the bottom chamber. After incubating at 37 °C with 5% CO2 for 24 h, cells were fixed with 4% paraformaldehyde and stained with 0.1% crystal violet solution. The cells that passed through the filter were photographed and counted by inverted fluorescence microscopy (Leica Microsystems GmbH, Wetzlar, Germany) in five randomly selected fields.
Wound healing assay
Cells were cultured in 6-well plates for 24 h, and a sterile pipette tip (200 μL) was used to scratch the monolayer. The cells were washed with PBS three times and were incubated in serum-free DMEM. Cell migration was photographed 24 h after scratching by an inverted microscope (Olympus, Japan), and the total wound area was analyzed using ImageJ software to assess the cell migration capacity.
Mouse xenograft model
Six-week-old female nude mice were purchased from the Laboratory Animal Center of Nanjing Medical University and maintained under pathogen-free conditions. Nude mice were subcutaneously injected in the inguinal region with 5× 106 GC cells (5 mice per group). Tumor volumes were measured every 4 days and calculated using the following formula: volume = length × (width/2)2. After 28 days, all the mice were sacrificed, and then tumors were resected and collected. To test chemosensitivity, one week after injection, cisplatin (5 mg/kg) in PBS was intraperitoneally injected into the mice three times per week. The xenograft tumors were harvested after 4 weeks. All experiments were approved by the Ethics Committee of the First Affiliated Hospital of Nanjing Medical University.
Luciferase reporter assay
GC cells were seeded in 96-well plates and then cotransfected with circCUL2/ROCK2 wild-type or mutant plasmids and miR-142-3p mimics or miR-NC using Lipofectamine 2000. After 48 h of incubation, the luciferase activity was measured according to the manufacturer’s instructions (Promega, Madison, WI, USA). Each experiment was repeated at least three times.
RNA immunoprecipitation (RIP)
The RIP assay was carried out by using a Magna RIP RNA Binding Protein Immunoprecipitation Kit (Millipore) according to the manufacturer’s instructions. The antibodies against AGO2 and IgG used for the RIP assays were purchased from Abcam (ab5072, rabbit polyclonal antibody, Cambridge, MA, USA).
RNA pull-down assay
The biotinylated circCUL2 probes were designed and synthesized by GenePharma (Shanghai, China). The circCUL2 probe was first incubated with C-1 magnetic beads (Life Technologies, Waltham, MA, USA) for 2 h at 25 °C to coat the beads with the probe. After the cells were harvested and lysed, the lysates were incubated with the circCUL2 probe or oligo probe at 4 °C overnight. The RNA-beads complexes were extracted with a RNeasy Mini Kit. The abundances of circCUL2 and miR-142-3p were evaluated by qRT-PCR.
Apoptosis analysis
Based on the manufacturer’s guidelines, cell apoptosis was measured by a flow cytometer (FACSCalibur, BD, USA). After 24 h of treatment, the cells were washed, resuspended, and then stained with FITC and PI, after which flow cytometry was used to analyze the apoptosis rate of cells treated under different conditions. The cell apoptosis data were analyzed by FlowJo V10 software (Tree Star, San Francisco, CA, USA). Each experiment was performed more than three times.
Cell viability analysis
Cells with or without transfection were seeded in 96-well plates at 5 × 103 cells/well for 24 h and then treated with cisplatin (DDP; 0, 1, 2.5, 5, 7.5, 15, 25 μg/mL), 5-fluorouracil (5-FU; 0, 1.5, 3, 6, 12, 25, and 45 μg/mL), doxorubicin hydrochloride (DOX; 0, 0.1, 0.5, 1, 2.5, 5, and 10 μg/mL), and mitomycin C (MMC; 0, 0.1, 0.5, 1, 2.5, 5, and 10 μg/mL) for 48 h. The cell viability was detected using an MTT Assay Kit (Sigma, St. Louis, MO, USA) to measure the cytotoxicity of cisplatin. The half maximal inhibitory concentration (IC50) was determined based on the absorbance at 490 nm as measured in a microplate reader (Bio-Rad, CA, USA).
Immunofluorescence
Cells in confocal dishes were treated for 48 h, fixed with 4% paraformaldehyde, treated with Triton X-100, blocked with goat serum, incubated with a primary antibody (against LC3 (1:100, #12741, Cell Signaling Technology)) and a secondary antibody, stained with DAPI, and viewed using a confocal laser scanning microscope (ZEISS, Germany).
Transmission electron microscopy (TEM)
Cells were fixed in 2% glutaraldehyde overnight at 4 °C, incubated in 1% osmium tetroxide for 1 h at 4 °C, dehydrated in graded ethanol, saturated in graded acetone, cut into 50-nm ultrathin sections, stained with lead citrate and viewed using JEM-1010 TEM (JEOL, Japan).
Statistical analysis
Data are reported as the means ± standard deviations (SDs), and GraphPad Prism (version 7.0) was used for the general statistical analysis. Student’s t-test was used to analyze the differences between two groups. The chi-square test was used to investigate the significance of the correlation of circCUL2 expression with the clinicopathological features. The survival curves were calculated by the Kaplan-Meier method and compared with the log-rank test. A P-value less than 0.05 was considered statistically significant.
Discussion
Recently, circRNAs have been proven to be differentially expressed in various diseases, especially cancers, due to their stability, abundance, conservation, and spatiotemporal specificity [
22‐
25]. Currently, many studies have detected the differential expression of circRNAs in GC tissues [
26‐
28]. In this study, we found that circCUL2 was downregulated in both GC tissues and GC cell lines, which suggested that circCUL2 may act as a factor to regulate GC progression. In addition, circRNAs are involved in all kinds of physiological and pathological processes, including proliferation, invasion, migration and chemoresistance, in GC development and treatment [
29,
30]. Here, we revealed that circCUL2 overexpression inhibited GC cell proliferation, migration, invasion and chemoresistance in vitro. Xenograft experiments confirmed these results in vivo. However, in hepatocellular carcinoma, circCUL2 was found to promote epithelial-mesenchymal transition (EMT), tumor metastasis and malignancy [
31]. The difference in expression and function of circCUL2 may be related to the tissue specificity of circRNA.
Accumulating evidence indicates that circRNAs could be miRNA sponges [
7,
32]. circRNAs always play functional roles as tumor regulators by binding to miRNAs as competing endogenous RNAs (ceRNAs), therefore modulating miRNA target gene expression. circRHOBTB3 might function as a ceRNA for miR-654-3p to inhibit the growth of GC by activating the p21 signaling pathway [
33]. In our study, miR-142-3p was selected as the candidate target miRNA of circCUL2 through bioinformatics analysis. A study integrating 4 GEO datasets (GSE93415, GSE23739, GSE28700, and GSE26645) showed that miR-142-3p was significantly upregulated in GC tissues [
34]. In addition, miR-142-3p expression in gastric MALT lymphoma was significantly increased compared with that in chronic gastritis, consistent with the results of this study [
35]. However, the role of miR-142-3p in tumors is still unclear, as miR-142-3p was downregulated in GC samples in some studies [
36,
37]. In prostate cancer, reducing miR-142-3p expression can significantly inhibit cell proliferation and induce cell cycle arrest [
38]. Herein, miR-142-3p was upregulated in GC tissues and promoted tumor metastasis and malignancy. The GC samples used in different studies may have different origins or differ in tissue structure. Furthermore, the miR-142-3p may act on different targets and play different roles in GC cells with different degrees of differentiation. FISH assays were then conducted to verify the colocalization in the cytoplasm and opposite expression of circCUL2 and miR-142-3p. In addition, luciferase reporter, RNA pull-down and RIP assays confirmed the direct interaction of circCUL2 and miR-142-3p. Furthermore, ROCK2, a serine-threonine kinase that can act on the cytoskeleton to regulate the morphology and migration ability of cells [
39], was predicted as a miR-142-3p target gene and was confirmed by dual-luciferase reporter assays. In colorectal cancer, the inhibition of ROCK2 expression triggers the initial polarization of the colon cancer cell line and induces cell invasion [
40]. In this study, ROCK2 was downregulated in GC tissues, and the correlation of circCUL2, miR-142-3p and ROCK2 expression was also tested in vivo and in vitro.
Cisplatin treatment is one of the most predominant chemotherapeutic strategies for patients with GC [
41], and chemoresistance is the main reason leading to poor prognosis of GC patients [
42]. In recent studies, circRNA plays a vital role in the regulation of GC cell cisplatin sensitivity. circFN1, which is increased in cisplatin-resistant GC tissues and cells, promotes the proliferation and inhibits the apoptosis of GC cells exposed to cisplatin in vivo and vitro [
43]. Our study showed that circCUL2 regulated cisplatin sensitivity both in vivo and in vitro. Analysis of the TCGA database showed that expression of miR-142-3p and ROCK2 was significantly correlated with the survival time of GC patients after chemotherapy, which suggested that miR-142-3p and ROCK2 may also be involved in chemotherapeutic resistance in GC. Studies have analyzed the miRNA profiles of chemo-sensitive and chemo-resistant GC samples and found that miR-142-3p was significantly upregulated in chemotherapy-resistant GC patients [
17]. In breast cancer cells, the downregulation of ROCK2 changed the hardness of the extracellular matrix, which affects the chemosensitivity of tumor cells [
44]. However, in pancreatic cancer and colon cancer, ROCK2 can promote the resistance of tumor cells to chemotherapy [
45,
46]. This research revealed that circCUL2 may affect the cisplatin sensitivity of GC cells by regulating miR-142-3p/ROCK2 by cotransfecting circCUL2 overexpression plasmids and miR-142-3p mimics.
Autophagy plays an important role in the mechanism of chemoresistance in GC cells. Abnormally activated autophagy induced by chemotherapeutic drugs could provide energy to support cancer cells, thereby promoting chemotherapeutic resistance [
47]. Previous studies have shown that circRNA regulates downstream target genes through the mechanism of ceRNA to mediate autophagy and participate in the regulation of drug resistance in GC cells [
48]. In addition, during cardiac fibrosis, ROCK2 knockout promotes age-related or starvation-induced autophagy activation [
21]. Moreover, ROCK2 inhibition can promote autophagy and reduce hippocampal damage during subarachnoid hemorrhage (SAH) [
49]. Intriguingly, the results of this study showed that ROCK2 was significantly increased in cisplatin-resistant cells overexpressing circCUL2, and ROCK2 might be the key mechanism by which circCUL2 regulates autophagy and drug resistance in GC cells through miR-142-3p.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.