Clinical specimens

Clinical specimens of CRC tumor samples and para-cancerous normal specimens (n = 76 pairs) were obtained from CRC patients undergoing surgery at the Second Hospital of Hebei Medical University between January 2012 and October 2014. The study protocol was approved by the Research Ethics Committee of the Second Hospital of Hebei Medical University. Patients provided informed consent for the collection of their clinical specimens. Notably, all the enrolled patients with CRC had not received radiation or chemotherapy prior to the surgery, and had not been diagnosed with other types of malignancies. Following the surgical resection, the specimens were pathologically verified and promptly stored in liquid nitrogen. The stage categorization of CRC patients were based on the American Joint Committee on Cancer's TNM staging method (7th edition). Table 1 presents a breakdown of clinical and pathological information for the CRC patients based on the expression levels of circCOL1A1 in the tumor specimens. The median expression level of circCOL1A1 in the tumor samples of 76 CRC patients was used as the cut-off to assign the patients into high-expression and low-expression groups.

Table 1 Correlation of circCOL1A1 expression with clinicopathologic features of colorectal cancerCell lines and culture specifications

The human CRC cell lines (HCT116, SW620, SW480, HCT8, and LoVo) and normal colonic epithelial cell line (FHC) were obtained from the Chinese Academy of Science Cell Bank (Shanghai, China) and the FHC human normal colorectal epithelial cell line was sourced from the American Type Culture Collection (ATCC, Manassas, VA, USA). All cell lines were cultured in the appropriate basal medium supplemented with FBS (Gibco, Waltham, MA, USA) and 1% antibiotic/antimycotic solution, then incubated at 37 °C with 5% CO2 in a humidified atmosphere. FHC and HCT116 cells were cultured in McCoy's 5A medium (KeyGEN, Nanjing, China), LoVo, and HCT8 cells in DMEM, and SW620 and SW480 cells in L-15 media (KeyGEN, Nanjing, China).

Lentivirus transduction for stable circCOL1A1 knockdown

Lentivirus particles were produced in 293T cells by transfecting the cells with the pLenti-puro vector carrying circCOL1A1 shRNA or control shRNA (GenePharma, Shanghai, China), together with psPAX2 and pCMV-VSV-G packaging plasmids. The supernatant containing lentiviral particles were then collected after 48 h of transfection. The transduction of HCT116 and SW480 cells was conducted using viral supernatant in the culture medium at the volume ratio of 2:1 in the presence of 8 μg/ml polyberne. 48 h after transduction, cells were selected with 800 ng/ml puromycin for 2 weeks, and the emerging cell clones with successful viral transduction were expanded and stored for further analysis.

Subcellular fractionation

The Cytoplasmic and Nuclear RNA Purification Kit (Invitrogen, Waltham, MA, USA) was utilized to isolate nuclear and cytoplasmic fractions, and the total RNA in each fraction was purified using Trizol reagent (Beyotime, Beijing, China) based on the supplier's instructions. Subsequently, the presence of circCOL1A1 in the cytoplasmic and nuclear compartment was assessed through RT-qPCR analysis, with glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as the marker for the cytoplasmic fraction and U6 small nuclear RNA (snRNA) for the nuclear compartment.

Invasion, migration, proliferation, and colony formation assays

The migration and invasion potentials of cells were evaluated via the Transwell inserts (Corning Incorporated, Corning, USA). It is important to mention that, while the protocols for assessing migration and invasion were similar, the key distinction is that the inserts were pre-coated with Matrigel (BD Biosciences, Franklin Lakes, NY, USA) for invasion assays but not necessary for migration. In brief, 2 × 105 cells were seeded into Transwell upper inserts using serum-free media and the lower compartment was filled with media supplemented with 10% FBS. Cells were cultured at 37 °C for 24 h. Subsequently, migrating and invading cells on the insert were fixed and then stained with 0.1% crystal violet (Vicmed, Shenzhen, China) for 15 min. The images were captured at 100× magnification using an Olympus microscope (Olympus, Tokyo, Japan). Finally, ImageJ software was utilized to quantify the number of invading cells. Each experiment was repeated three times.

Cell proliferation was assessed using the Cell Counting Kit (CCK)-8 assay (Sigma-Aldrich, St. Louis, USA). Initially, cells were seeded at a density of 5 × 103 cells/well in a 96-well plate. At indicated time point, 10 μl of CCK-8 solution was added to each well and the plate was further incubated at 37 °C for two hours. The optical density (OD) was measured at 450 nm to quantify cell proliferation capacity. Similarly, the cells were seeded at 2 × 103 cells per well in a 6-well plate and grown in media with 10% FBS for 14 days for the colony formation assay. The medium was changed every four days. The colonies formed were fixed in methanol and stained with 0.1% crystal violet for 15 min. The number of stained colonies was counted to determine the rate of colony formation. All experiment was performed three times independently.

GLS1 expression vector construction and transfection

The cDNA sequence of human GLS1 gene was synthesized and inserted into the expression vector (pcDNA3.1) by the Shanghai Genomeditech Co. Ltd. (Shanghai, China). Synthetic miRNA mimic, inhibitor and corresponding controls (miR-NC and inhibitor control) were produced by RiboBio (Guangzhou, China). For transfection, cells were firstly seeded in a 6-well plate at a density of 1 × 106 cells/well and incubated until 70% confluence. Cells were then transfected with 200 nm miRNA mimic/inhibitor or 6 μg expression vector using lipofectamine® 2000 (Thermo Fisher Scientific, Inc., Waltham, USA) according to the manufacturer's instructions. Transfected cells were harvested after 72 h for further functional assay and molecular analysis.

RT-qPCR analysis

As per the manufacturer's guidelines, RNA was isolated from specific cell lines utilizing the Roche Isolation Reagent (Roche, Basel, Switzerland). The generation of complementary DNA (cDNA) was carried out using the Moloney Murine Leukemia Virus reverse transcriptase kit (MMLV RT Kit, ABI, Warrington, UK). RT-qPCR was conducted using the SYBR PCR Kit (Roche) on the ABI 7500 qPCR quantification system. Subsequently, gene expression levels were determined relative to the control gene GAPDH using the delta–delta Ct method. Provided below are the detailed sequences of primers:

circCOL1A1: Forward—CAAGGGTCTGACTGGAAGCC

Reverse – CGAGCTCCTCGCTTTCCTTC

miR-214-3p: Forward—GCGACAGCAGGCACAGACA

Reverse-AGTGCAGGGTCCGAGGTATT

GAPDH: Forward—GGAGCGAGATCCCTCCAAAAT

Reverse—GGCTGTTGTCATACTTCTCATGG

GLS1: Forward—AGTTGCTGGGGGCATTCTTTTAGTT

Reverse—CCTTTGATCACCACCTTCTCTTCGA

Western blotting

The RIPA lysis buffer (Beyotime, Beijing, China) was used to collect protein samples from cultured cells at 4 °C for 10 min. Subsequently, the mixture was centrifuged at 14,000 g for 10 min at 4 °C. The supernatant was collected for protein measurement by a Bradford assay kit (Zeye Biotech, Shanghai, China). Next, 20 µg of denatured protein sample was loaded into each lane of a 10% sodium dodecyl sulfate–polyacrylamide gel for electrophoresis, followed by the transfer onto a polyvinylidene difluoride (PVDF) membrane. The membrane was then blocked with 5% non-fat milk before incubating overnight at 4 °C with primary antibodies for GLS1 and GADPH (Invitrogen, 1:1000). The next day, the membrane was further probed with horseradish peroxidase (HRP)-linked secondary antibodies (1: 5000, Abcam, Cambridge, UK) for 1 h at room temperature. Protein bands were visualized using BeyoECL PLUS Supersensitive ECL Chemiluminescence Kit (Beyotime, Beijing, China).

RNA immunoprecipitation (RIP) assay

A Magna RNA-binding protein immunoprecipitation kit (Millipore, Billerica, MA, USA) was used to carry out the RIP experiment, according to the manufacturer’s instructions. Briefly, 10 μg anti-Ago2 or IgG antibodies (Abcam, Cambridge, UK) were immobilized with the protein A/G beads and then incubated with the cell extracts from 1 million cells for 4 h at °C. After washing, the RNA samples on the beads were purified using Trizol reagent and RT-qPCR analysis was conducted to quantify the precipitated RNAs.

Dual luciferase reporter assay

The predicted interacting sites (WT) or the mutated sequences (MUT) were cloned into the pmirGLO firefly luciferase reporter vector (Promega Corp., Madison, USA). The cells at a logarithmic phase were seeded in the 24-well plates at a density of 1.5 × 104/well. 1 μg WT or MUT reporter vector and 1 μg pRL Renilla control vector were introducted into CRC cells using the LipofectamineTM 2000 reagent. After 48 h of transfection, a dual-luciferase reporter assay system (Promega) was utilized to detect firefly luciferase activity from the reporter vector and renilla luciferase activity from the control vector, respectively.

RNA pull-down experiment

The lysates of CRC cells (1 × 106 cells) were collected by IP lysis buffer (Beyotime) and were incubated with 200 nM biotinylated circCOL1A1 probe and control oligos for 2 h at 4 °C. 100 μL M-280 streptavidin magnetic beads (Sigma-Aldrich) was added to the mixture for overnight incubation. A magnetic rod was employed to bring down the magnetic spheres and linked nucleic acids, following which the specimens underwent a total of 4 washes with a high-salt washing solution. The RNA samples from the magnetic pull-down were extracted using Trizol reagent and were subjected to RT-qPCR analysis.

Immunohistochemical staining

Initially, CRC samples were fixed and encased in paraffin, then cut into 4-μm slices. Following this, the samples were deparaffinized and rehydrated. The activity of peroxidase was blocked by exposing the slides to 95 °C in citrate buffer (0.01 M, pH 6.0) for 20 min. Next, goat serum was applied to the slices for 1 h and the sample was further incubated overnight at 4 °C with anti-Ki-67 antibody (1:500, Abcam, Cambridge, UK). Next day, after the labeling with (HRP)-linked secondary antibody, the sample was treated with 3,3ʹ-diaminobenzidine substrate (DAB; Zhongshan Biotech, Beijing, China) for 5 min for color development. Then, all sections were dehydrated, stained with hematoxylin, and sealed before image capturing under Olympus BX-51 light microscope.

Detection of glutamine consumption and glutamate

CRC cells with different treatment were grown in the basal medium supplemented with glutamine (2 mM). After 24 h of incubation, the intracellular glutamine and glutamate levels were determined using the corresponding assay kits (Biovision, Milpitas, CA, USA) following the manufacturer's instructions.

Bioinformatics analysis

The interactions between circCOL1A1 and miRNAs were predicted using the starBase (http://starbase.info/) and circBANK (http://www.circbank.cn/) databases. The potential target miRNAs of GLS mRNA 3's untranslated region (UTR) was identified using the starBase database.

Tumor xenograft experiments

All animal experiments were approved by the Animal Ethics Committee of Second Hospital of Hebei Medical University. Notably, all animal procedures were performed following the National Institutes of Health's Guide for the Care and Use of Laboratory Animals. Female BALB/c nude mice aged four to six weeks were provided by the HFK Bioscience (Beijing, China). Animals were randomly distributed into the control and circCOL1A1 silencing groups (n = 6 animals in each group). To set up the xenograft, 5 × 106 sh-NC HCT116 cells or sh-circCOL1A1 HCT116 cells were injected subcutaneously in the animal along with 200 μl of Matrigel. The formation of xenograft tumor was monitored every week. After 5 weeks, all the mice receiving subcutaneous injections were sacrificed by cervical dislocation. The xenograft tumor samples were harvested for further analysis.

Statistical analysis

Statistical analysis was conducted using SPSS 17.0 software (SPSS, USA), with GraphPad Prism 7 (La Jolla, USA) being utilized for data visualization. The data were presented as mean and standard deviation (S.D.). Two-tailed Student's t test or analysis of variance (ANOVA) was employed to identify significant group differences, with statistical significance set at P < 0.05. Association between the expression of two molecules was evaluated using Spearman's correlation analysis. Overall survival (OS) rate was calculated using the Kaplan–Meier method and log-rank test. (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001).