Introduction
Colorectal cancer (CRC) is one of the most common malignant neoplasms worldwide; the number of CRC cases increases every year, and CRC poses a serious threat to human life and health [
1]. The unknown etiology, lack of obvious symptoms in the early stages, and high level of metastasis are important factors leading to the dismal prognosis and high mortality for CRC patients [
2]. Although progress has been made in diagnostic and therapeutic strategies, the clinical outcomes and prognoses of CRC patients with advanced-stage disease remain poor [
3]. Thus, further research on the molecular mechanisms that regulate the progression and migration of CRC represents a crucial step in the exploration of novel molecular targets, which may help to generate more effective therapies.
Circular RNAs (circRNAs) are a class of RNA molecules that form single-stranded closed loop structures through covalent bonds without 5′ or 3′ free ends [
4]. Following the advent of high-throughput sequencing and computational approaches, thousands of circRNAs have been identified. The majority of circRNAs are exonic circRNAs, which are derived from exonic regions of known protein-coding genes by back-splicing [
5]. In recent years, an increasing number of studies have identified circRNAs that play crucial roles in tumor carcinogenesis by sponging microRNA (miRNAs) [
6]. For example, the well-known circRNA ciRS7 abrogates the tumor suppressive effect of miR-7 to promote the progression of esophageal squamous cell carcinoma [
7] and colorectal cancers [
8]. Circ_0039569 promotes renal cell carcinoma growth and metastasis by regulating miR-34a-5p/CCL22 [
9]. Circ-ZEB1.33 promotes the proliferation of human hepatocellular carcinoma (HCC) by sponging miR-200a-3p and upregulating the expression of CDK6 [
10].
MiR-340 has been reported as a tumor suppressor gene that can regulate the cell cycle and affect tumor migration and metastasis in several types of neoplasms, including glioblastoma multiforme [
11,
12], non-small cell lung cancer [
13], breast cancer [
14], ovarian cancer [
15] and gastric cancer [
16]. Furthermore, the expression of miR-340 in bone marrow negatively correlates with liver metastasis of CRC [
17]. Zhang et al. showed that mir-340 suppresses the growth and enhances the chemosensitivity of CRC by targeting RLIP76 [
18]. Sun et al. demonstrated that miR-340 inhibits the growth of CRC by neutralizing the Warburg effect by regulating the alternative splicing of the PKM gene [
19]. We used a bioinformatics website (
http://starbase.sysu.edu.cn/index.php) to predict that miR-340 has binding sequence within many circRNAs. We detected the expression of these circRNAs in miR-340 overexpressed cells and 20 fresh colon cancer tissues. The results showed that only hsa_circ_001680 had a negative correlation with miR-340. Therefore, we suspected potential interactions exist between circ_001680 and miR-340 and we focused our research on circ_001680.
Circ_001680 (circBase ID: hsa_circ_0000598), is a circRNA located at position of chr15:45009906–45,009,989. The gene symbol is B2M. It has not been reported in any tumors. The mechanism of miR-340 and circ_001680 in the progression of CRC has not been elucidated, and the specific function of circ_001680 in the development of CRC requires further study.
In this study, we demonstrated that circ_001680 could promote the proliferation and migration of CRC cells. Furthermore, circ_001680 was shown to inhibit the expression of miR-340 by acting as an RNA sponge. In addition, circ_001680 could upregulate the miR-340 target gene BMI1, promote the cancer stem cell (CSC) population of CRC cells and induce irinotecan chemotherapy resistance. Our results highlight a new molecular mechanism underlying the tumorigenicity of colon cancer cells and suggest that circ_001680 is a potential chemotherapy resistance marker in CRC.
Materials and methods
Tissue specimens and cell cultures
Forty-two pairs of freshly CRC specimens and their matched adjacent paracancerous normal colorectal tissues were recruited from the Department of General Surgery, Nanfang Hospital for histological analysis. All tissues were stored in liquid nitrogen for further use. The medical records of the patients were collected, and the following information was obtained: age, sex, pathological stage, T stage, lymph node metastases and distant metastasis.
Human colorectal cell lines (FHC, HCT116, SW480, HCT15, SW620, CACO2, DLD1, LOVO, HT29, HCT8 and RKO) were purchased from American Type Culture Collection Cell Biology Collection and were maintained in the Department of Pathology, Southern Medical University. Cells were cultured in RPMI 1640 (Invitrogen, Carlsbad, CA, USA) or DMEM (Invitrogen, Carlsbad, CA, USA) supplemented with 10% FBS (Invitrogen, Carlsbad, CA, USA) and 1% penicillin/streptomycin (Invitrogen, Carlsbad, CA, USA) at 37 °C with 5% CO2.
Wound healing assay
A total of 5 × 105 cells/well in the logarithmic growth phase were seeded into 6-well plates. When the cell density reached 80 to 90%, a scratch was made in the monolayer in the middle of the well with a 100 μl pipette tip. The tip was kept perpendicular to the bottom of the well to obtain a straight gap. The detached cells were washed away and removed three times a day. Wound healing within the same scraped line was then observed and photographed at the indicated time points (0 h, 24 h, 48 h, and 96 h). Each experiment was repeated three times.
RNA extraction and qRT-PCR
Total tissue mRNA was extracted from tissues with a mirVana miRNA Isolation Kit (Ambion, Austin, TX, USA) according to the manufacturer’s protocol. Then, we synthesized cDNA from total RNA using the TaqMan miRNA Reverse Transcription Kit (Applied Biosystems, Foster City, CA, USA). qRT-PCR was performed on the Applied Biosystems 7500 Sequence Detection system with IQTM SYBR Green Supermix (Bio-Rad Laboratories, Hercules, CA, USA) and 5 ng of cDNA and 10 pM of each primer. The cycling conditions were set as previously described [
20]. The data were normalized to the geometric mean of the housekeeping gene GAPDH or to U6 small nuclear RNA expression and calculated according to the 2
-ΔΔCT method. The forward and reverse primer sequences are shown in Additional file
1: Table S1.
Migration assay
A total of 1 × 104 cells were seeded into the upper Boyden chamber with an 8 μm pore size filter membrane and culture medium supplemented with 10% fetal bovine serum was added to the lower chamber as a chemoattractant. Twenty-four hours later, the cells on the upper filter were gently removed with a cotton swab. The cells that had migrated to the lower surface of the filter were fixed in 4% paraformaldehyde and stained with hematoxylin for 10 min. Then, the cells were washed three times to remove the hematoxylin. The filter membrane was dried with a blower, and the migrated cells (10 random 200× fields per well) were counted. Three independent experiments were performed and the data are presented as the mean ± s.e.m.
Detection CRC cells were digested and seeded directly in 6-well plates (5 × 102 cells/well) for the colony formation assay and cultured in the presence of 10% FBS and 1% penicillin/streptomycin at 37 °C with 5% CO2. Two weeks later, the medium was removed, and the plates were washed with phosphate-buffered saline (PBS) three times. The cells were fixed with anhydrous ethanol for 30 min and then stained with hematoxylin for 20 min. The plates were dried with a blower to ensure that high-quality images were obtained. The colonies were defined as > 50 cells/colony.
Xenograft model in nude mice
For tumourigenesis assays, 2 × 10
6 cells per mouse were subcutaneously injected into the right dorsal flanks of female BALB/c athymic nude mice (4–6 weeks of age, 18–20 g), which were obtained from the Animal Center of Southern Medical University, Guangzhou, China). The mice were sacrificed at approximately 7 to 8 weeks. The tumors were excised and placed in 10% neutral buffered formalin for 24 h. The tumors were excised and placed in 10% neutral buffered formalin for 24 h [
21]. All mice were raised under specific pathogen-free conditions, and all experiments were approved by the Use Committee for Animal Care and were performed in accordance with institutional guidelines. The tumor size was measured using a slide caliper, and the tumor volume was determined by the following formula: 0.44 × A × B
2, where A represents the diameter of the base of the tumor and B represents the corresponding perpendicular value.
Immunohistochemistry
The tissues were cut into 4 μm-thick sections, baked at 65 °C for 30 min, and then dewaxed by dimethylbenzene and alcohol. The sections were then deparaffinized with xylene and treated with 3% hydrogen peroxide to attenuate endogenous peroxidase activity. Next, the sections were submerged in citrate buffer for antigen retrieval and incubated with 1% bovine serum albumin (BSA) to block nonspecific binding. Primary antibodies against Ki67 (1:500; ZSGB-BIO, Beijing, China), caspase-3 (1:500; ZSGB-BIO, Beijing, China), CD133 (1:200; Cell Signaling Technology, Danvers, MA, USA), BMI1 (1:200; Cell Signaling Technology, Danvers, MA, USA), and SOX-2 (1:200; Cell Signaling Technology, Danvers, MA, USA) and an appropriate secondary antibody (1:500, ZSGB-BIO, Beijing, China) were used according to the manufacturer’s instructions. The sections were incubated with DAB and hematoxylin and then scored independently by two observers. The score was based on both the proportion of positively stained tumor cells and the intensity of staining.
Luciferase assays
CRC cells were seeded in triplicate into 24-well plates (1 × 105 cells per well) and then cultured for 24 h. The constructed pGL3-basic luciferase reporter plasmid (1.5 μg, Promega) or the control luciferase plasmid (1.5 μg, Promega) was cotransfected into the cells with the pRL-SV40 plasmid (0.15 μg, Promega) using Lipofectamine 2000 Reagent (Invitrogen). Luciferase and Renilla activities were detected 36 h after transfection using the Dual-Luciferase Reporter Assay Kit (Promega) according to the manufacturer’s protocol. All experiments were conducted at least three times, and the data are presented as the mean ± SD.
Flow cytometry assay
The indicated CRC cells were incubated with 5 μg/ml irinotecan (Selleck, S2217) or DMSO for 36 h and then placed in a 1.5 ml tube. The cells were then washed twice with PBS and centrifuged at 400 x g for 2 min. The supernatant was discarded, and the cells were resuspended in 60 μl of surface staining buffer (PBS, pH 7.4, 0.1% BSA) containing antibodies against CD44 and CD133 at 1 μg/ml (BD Pharmingen, Franklin Lakes, NJ, USA) and incubated for 30 min at 4 °C. Then, the cells were resuspended in PBS without washing and analyzed on a FACS flow cytometer according to the manufacturer’s instructions. The results were analyzed by FlowJo software.
Western blot analysis
Western blotting was performed according to a previous study [
22]. Protein lysates were prepared, subjected to SDS-PAGE, transferred onto PVDF membranes and blotted according to standard methods using anti-BMI1 (Cell Signaling Technology, Danvers, MA, USA), anti-CD44 (BD Pharmingen, Franklin Lakes, NJ, USA), anti-CD133 (BD Pharmingen, Franklin Lakes, NJ, USA), and anti-SOX2 (Cell Signaling Technology, Danvers, MA, USA); an anti-α-tubulin monoclonal antibody (Sigma, St Louis, MO, USA) served as a loading control.
Fluorescence in situ hybridization (FISH)
The digoxin-labeled probes specific to circ_001680 and biotin-labeled probes against miR-340 were prepared by Geneseed Biotech, and the sequences are showen in Additional file
1 Table S1. SW480 and HCT116 cells were cultured on coverslips and fixed with 4% paraformaldehyde in PBS for 15 min. The probes were diluted in hybridization solution (Geneseed Biotech, Guangzhou, China) in PCR tubes and were heated at 95 °C for 2 min in a PCR block to denature the probe. The probe was immediately chilled on ice to prevent reannealing. The hybridization solution was drained, and 100 μL of diluted probe per section was added to cover the entire sample. The samples were covered with a coverslip to prevent evaporation and were incubated in the humidified hybridization chamber at 65 °C overnight. The signals were detected by Cy3-conjugated anti-digoxin and FITC-conjugated anti-biotin antibodies (Jackson ImmunoResearch Inc., West Grove, PA, USA). Cell nuclei were counterstained with 4,6-diamidino-2-phenylindole (DAPI). Finally, the images were obtained on a Zeiss LSM 700 confocal microscope (Carl Zeiss, Oberkochen, Germany) [
23‐
26].
RNA pull-down assays
RNA pull-down assays were performed with PierceTM Magnetic RNA-Protein Pull-Down Kit (Millipore, Billerica, MA, USA) following the manufacturer’s suggestions, biotinylated circ_001680, biotinylated BMI1 (Geneseed, Guangzhou, China, the sequence is shown in Additional file
1: Table S1) or biotinylated negative control (NC) was incubated with the RIP lysates from SW480 and HCT116 cells for 2 h at 25 °C. The circ_001680/miR-340 or BMI1/miR-340 complexes were captured with Streptavidin-coupled Dynabeads for 1 h at 25 °C. Then the complexes were incubated with wash buffer containing proteinase K for 1 h at 25 °C. The complexes in the pull-down were determined using qRT-PCR analysis. All tests were carried out in triplicate [
23‐
26].
RNA immunoprecipitation
RNA immunopreciptiation (RIP) assay was conducted by an EZ-Magna RIP Kit (Millipore, Billerica, MA, USA) according to the manufacturer’s protocol. The AGO2-RIP experiments were performed in SW480 and HCT116 cells transiently overexpressing miR-340 or miR-NC. Forty-eight hours later, Approximately 1 × 10
7 cells were collected and dissolved in 100% RIP Lysis Buffer with proteinase and RNase inhibitors, and the RIP lysates were incubated with RIP buffer containing magnetic beads conjugated with human anti-Ago2 antibody or nonspecific mouse IgG antibody (Cell Signaling Technology, USA). Twenty-four hours later, the RNA/bead complex was washed five times and resuspended in buffer supplemented with RNase-free DNase and proteinase K. The immunoprecipitated RNAs were subjected to qRT-PCR to detect the enrichment [
23‐
26]. All tests were carried out in triplicate.
A mixture culture medium was prepared and included serum-free 1640 medium (Invitrogen), 2% B-27 Supplement (Invitrogen, Carlsbad, CA, USA), 20 ng/ml basal fibroblast growth factor (bFGF) (PeproTech, Rocky Hill, NJ, USA), 20 ng/ml epidermal growth factor (EGF) (PeproTech), 0.4% BSA (Sigma-Aldrich), and 5 μg/ml insulin (Sigma-Aldrich). CRC cells were digested and resuspended in the prepared medium. Cells (1 × 103 cells per well) were seeded in 6-well ultralow attachment plates. The cells were grown in the prepared medium in an incubator at 37 °C and 5% CO2 with saturated humidity for 12–14 days. The tumor sphere was defined as > 2000 cells. Sphere efficiency was defined as the percentage of the number of spheres divided by the original number of seeded cells. All experiments were conducted at least three times, and the data are presented as the mean ± SD.
Irinotecan treatment experiment
For the tumor drug treatment assay, 2 × 106 cells per mouse were subcutaneously injected into the right dorsal flanks of female BALB/c athymic nude mice (4–6 weeks of age, 18–20 g), which were obtained from the Animal Center of Southern Medical University, Guangzhou, China). The tumor-bearing mice were observed until the tumor volume reached approximately 150 mm3, and the mice were randomly grouped. Mice in the two groups were intraperitoneally injected with 20 mg/kg irinotecan (Selleck, S2217) or DMSO three times per week. The tumor diameters were measured twice a week. Approximately 50 days later, the tumors were excised and placed in 10% neutral buffered formalin for 24 h. For flow cytometry, the indicated cells were seeded into plates with the medium as mentioned above, and 5 μg/ml irinotecan (Selleck, S2217) or PBS was added when the cells adhered to the plates. After 48 h, the cells were detected by flow cytometry according to the manufacturer’s instructions. For tumor sphere formation analysis, a mixture culture medium was prepared as mentioned above and an additional 5 μg/ml irinotecan (Selleck, S2217) or PBS was added. Then, the cells were cultured in these medium in an incubator at 37 °C and 5% CO2 with saturated humidity for 12–14 days.
Statistical analyses
All data were plotted and counted by SPSS19.0 for Windows, represented the mean ± SD. P < 0.05 was considered to be statistically significant. The difference in the miR-340 or hsa_circ_001680 expression level between carcinomatous and normal CRC tissues was evaluated by a paired t test. Clinical pathological characteristics of circ_001680 expression in CRC patients were analyzed by a two-sample t test. The linear relationship between circ_001680 and miR-340 expression levels in colorectal cancer cells was measured by Pearson correlation coefficient.
Discussion
In recent years, an increasing number of studies have identified circRNAs that may serve as diagnostic or predictive biomarkers of some diseases, especially cancers [
37,
38]. For instance, hsa_circ_0013958 may be useful as a potential noninvasive biomarker for the early detection and screening of lung adenocarcinoma (LAC) [
39]. Circ_0026344 acts as a prognostic biomarker to suppress colorectal cancer progression via microRNA-21 and microRNA-31 [
40]. CircPVT1 is significantly overexpressed in osteosarcoma (OS) tissues, serum, and chemoresistant cell lines, suggesting that this circRNA is a potential diagnostic biomarker with useful sensitivity and specificity [
41].
In our research, functional experiments showed that circ_001680 overexpressing cell lines could increase the cellular capabilities of proliferation and migration. At the tissue level, we used qRT-PCR to assess the expression of circ_001680 in human CRC tissues and their matched normal tissues, revealing that circ_001680 was expressed at higher levels in CRC tissues than in their matched normal tissues. Bioinformatics predictions and the dual-luciferase reporter experiments demonstrated that circ_001680 could target miR-340, which has been identified as a miRNA that is downregulated in many types of cancer. MiR-340 can inhibit cell proliferation, induce cell apoptosis, and reduce cell migration and invasion by targeting multiple oncogenes in breast cancer, gastric cancer, glioblastoma, and non-small-cell lung cancer. Bioinformatics predictions and the luciferase reporter system results showed that miR-340 could target the 3’UTR of BMI1 and that ectopic upregulation of BMI1 partially reversed the influence of miR-340 on CRC cell growth and migration (Additional file
4: Figure S3).
BMI1 is a core component of the polycomb repressive complex that mediates gene silencing via monoubiquitination of histone H2A [
42,
43]. Furthermore, BMI1 is an important stem cell self-renewal factor [
44,
45] that has been shown to be abnormally expressed in CRC and is associated with the self-renewal of CSCs in CRC [
46]. Targeting BMI1
+ CSCs has been shown to overcome chemoresistance and inhibit metastases in squamous cell carcinoma [
47] and gastric cancer [
32]. Liu et al. found that overexpression of miR-128-3p could reestablish sensitivity in resistant cells by reducing BMI1 expression, which is related to oxaliplatin resistance [
48]. CSCs are believed to function as a type of stem cell-like cell population in tumors that promote self-renewal [
49], and they are associated with tumor metastasis [
50,
51] and drug resistance [
34,
52]. In our research, we found that BMI1 could induce irinotecan chemotherapy resistance in CRC cells. We transiently transfected cells with BMI1 at different concentrations in the cells, and the results showed that higher concentrations of BMI1 were correlated with more obvious chemotherapy resistance effects (Additional file
5: Figure S4). Although irinotecan is known to be effective for patients with advanced CRC, chemotherapy resistance to the drug often leads to cancer treatment failure. It is necessary to investigate the biological basis of chemotherapy resistance and identify an effective marker.
Thus, we suspected that circ_001680 could affect the stem cells characteristics of CRC and induce chemoresistance by upregulating the expression of BMI1. There were no significant changes in the number of stem cell spheres in the circ_001680 group after treatment with irinotecan, whereas a notable decrease was observed in the control group in response to irinotecan treatment. Similarly, the levels of stem cell markers and the populations of CD133+/CD44+ cells were remarkably decreased in the control group compared with the levels in the circ_001680 group after treatment with irinotecan. We subcutaneously injected the indicated CRC cells into female BALB/c athymic nude mice. When the tumor volume reached approximately 150 mm3, mice were intraperitoneally injected with 20 mg/kg irinotecan or DMSO three times per week. Interestingly, the tumor volumes and weights of the mice in the drug groups were notably smaller than those in the DMSO group. However, upregulation of circ_001680 did not result in significantly different volumes and weights in the drug and DMSO groups. These findings demonstrate a novel role for circ_001680 in the regulation of stem cell characteristics and chemoresistance and provide a molecular basis for targeting BMI1 to overcome irinotecan chemoresistance in colon cancer.
Our study demonstrated that circ_001680 could mediate CRC tumor growth and migration for the first time. At the same time, we found an essential link connecting circ_001680, miR-340 and BMI1 in CRC. Furthermore, circ_001680 was shown to promote the CSC population in CRC and induce irinotecan chemoresistance by upregulating the miR-340 target gene BMI1. Understanding the precise function and mechanism of circ_001680 in the progression of CRC will increase our understanding of CRC biology and may also allow the development of a novel irinotecan chemoresistance therapeutic strategy.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.