Background

Colorectal cancer (CRC) is the second and third most frequently diagnosed cancer in females and males worldwide, respectively [1]. Despite significant advancements in surgical techniques, chemotherapy, immunotherapy, and molecular-targeted treatments, the overall prognosis and 5-year survival rate of CRC patients remain poor, particularly for those with distant metastatic CRC [2, 3]. Thus, it is crucial to understand the potential mechanisms underlying CRC progression, which may contribute to the enlargement of the molecular pool for the diagnosis and therapy of CRC.

CircRNAs, a newly recognized type of noncoding RNAs (ncRNAs), are generally characterized by the covalently closed loop structure without a 3′-poly-A tail or 5′-cap [

Methods

Microarray and RNA sequencing data

Two circRNA expression profiles (GSE121895 and GSE142837) and a miRNA expression profile (GSE115513) were downloaded from the Gene Expression Omnibus (GEO) database. The CRC RNA sequencing data (COAD, READ) were obtained from The Cancer Genome Atlas (TCGA) database. The “limma” package in R software was used to explore the significantly differentially expressed circRNAs or miRNAs.

Tissue specimens and ethical approval

One hundred pairs of fresh human CRC specimens and their adjacent normal tissues were obtained from Wuhan University’s Zhongnan Hospital. The study participants provided their informed consent before the study began. The clinical information and pathological features of these patients were collected and are presented in Supplementary Table S1.

Cell culture

The CRC cell lines (SW620, HCT116, SW480, DLD-1, LOVO, and HT29) and the human embryonic kidney epithelial cell line (HEK-293T) were purchased from the American Type Culture Collection (Manassas, VA, USA). HCT-116, DLD-1, HT29, and LOVO were grown in RPMI-1640 medium (Gibco; Thermo Fisher Scientific, Waltham, MA, USA) with 10% fetal bovine serum (FBS) (HyClone Laboratories Inc, Logan, UT, USA). SW480, SW620, and HEK-293T were cultured in Dulbecco’s Modified Eagle Medium (DMEM) (Gibco; Thermo Fisher Scientific) containing 10% FBS.

RNA extraction and reverse transcription quantitative PCR (RT-qPCR)

Total RNA from cells or CRC tissues was extracted using TRIzol reagent (Invitrogen, Waltham, MA, USA). cDNA was synthesized with random primers using HiScript II Q RT SuperMix (Vazyme, Nan**g, China) or miRNA stem-loop RT primers using a miRNA 1st Strand cDNA Synthesis Kit (Vazyme). Subsequently, RT-qPCR was performed using ChamQ Universal SYBR qPCR Master Mix (Vazyme). The primer sequences are listed in Supplementary Table S2 and were synthesized by Tsingke (Nan**g, China).

RNase R treatment assay and actinomycin D assay

For the RNase R degradation assay, total RNA (2 µg) was administrated with 3 U/µg RNase R (Epicenter). For the actinomycin D assay, SW480 and HT29 cells were cultured with 5 mg/L actinomycin D (Sigma-Aldrich, St. Louis, MO, USA) and incubated for indicated periods. Thereafter, RNA was extracted at each time point and subjected to RT-qPCR.

Western blot

The CRC tissue samples or cells were lysed and denatured in loading buffer. Subsequently, the total proteins were resolved on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gel and then transferred onto polyvinylidene difluoride (PVDF) membranes. After blocking in 5% non-fat milk, the PVDF membranes were reacted with primary antibodies overnight and then hybridized with the corresponding secondary antibodies. Finally, the protein bands were visualized by chemiluminescence using an ECL system. All primary antibodies are listed in Supplementary Table S3.

Fluorescence in situ hybridization (FISH)

Specific probes for miR-663b and circ_0003215 were synthesized by GenePharma (Shanghai, China) and RiboBio (Guangzhou, China). All probe sequences are listed in Supplementary Table S4. Briefly, SW480 and HT29 cells were seeded in confocal dishes (Nest Scientific Inc, Woodbridge, NJ, USA), washed, fixed, and then permeabilized with 0.1% Triton X-100. For probe detection, CRC cells were pre-hybridized for 40 min and subsequently hybridized with the specific probe mix overnight at 37 °C. Thereafter, the CRC cells were stained with DAPI at 37 °C for 30 min after washing with a hybridization solution. Finally, the fluorescence images were acquired using a Dragonfly 200 High-Speed Confocal Microscope (Andor Technology Ltd, Belfast, UK).

miRNAs, siRNAs, plasmid, and transfection

MiR-663b mimics, inhibitors, and small interfering RNAs (siRNAs) targeting circ_0003215 were synthesized by RiboBio. The pLCDH-circ_0003215 overexpression vector was constructed by cloning the cDNA of circ_0003215 into the specific pLCDH-circ expression vector. The short hairpin RNA (shRNA) of DLG4 was constructed by cloning the DNA sequence targeting DLG4 into the pLKO.1 plasmid. All the sequences are listed in Supplementary Table S4.

Nuclear and cytoplasmic extraction

Cytoplasmic and nuclear fractions of CRC cells were separated using a Cytoplasmic and Nuclear RNA Purification kit (Norgen Biotek Corp, Thorold, ON, Canada). Briefly, CRC cells were harvested and lysed on ice for 15 min. After centrifugation, the nuclear and cytoplasmic fractions were collected separately. Afterward, the isolated RNAs in the nucleus and cytoplasm were extracted using TRIzol and subjected to RT-qPCR.

Cell counting kit-8 (CCK-8) assay

Cell proliferative ability was determined using a CCK-8 assay kit (Beyotime Biotech, Bei**g, China). Briefly, CRC cells were seeded in a 96-well plate and then cultured overnight. At 24, 48, 72, and 96 h, CRC cells with different transfections were supplemented with CCK-8 solution (10 μL/well) and cultured for an additional 2 h. Finally, the resulting optical density at 450 nm was measured using a microplate reader (Bio-Rad Laboratories, Hercules, CA, USA).

Transwell assay

The transwell assays were conducted using transwell chambers (Corning Life Sciences, Corning, NY, USA) and matrigel (BD Biosciences, Franklin Lakes, NJ, USA). For the transwell migration assay, approximately 5 × 104 transfected CRC cells suspended in 200 µL of medium without FBS were seeded in the upper chamber, while the lower chamber contained culture medium with 20% FBS as a chemoattractant. For cell invasion assay, cells were also seeded in the upper chamber precoated with matrigel. Next, the CRC cells with different transfections subjected to migration assay for 24 h and invasion assay for 48 h assay After removal of the cells remaining on the upper surface of the membrane, the cells migrated or invaded undersurface of the membrane were fixed. After staining with 0.1% crystal violet, the cells were observed and counted under a microscope.

Colony formation assay

For the colony formation assay, 5 × 102 transfected CRC cells were seeded into a six-well plate and cultured for 14 d. Afterward, the cell colonies were fixed with paraformaldehyde and stained with 0.1% crystal violet. Next, the plate was washed with PBS, and the total number of cell colonies of more than 50 cells was counted and photographed.

Scratch wound healing assay

Transfected CRC cells were seeded in 6-well culture plates in complete DMEM/RPMI-1640 media and grown to 80% confluence in a monolayer overnight. Next, scratch wounds were inflicted along the diameter of each well using a 100-µL pipette tip. Thereafter, cell debris and suspended cells were removed by extensive washing, and then culture medium with 1% FBS was added. Finally, wound healing was measured and photographed under a microscope at 0 and 24 h within the wound line and the healing area was calculated using Image J software (NIH, Bethesda, MD, USA).

5-Ethynyl-2′-deoxyuridine (EdU) assay

A BeyoClick™ EdU-594 Cell Proliferation Kit (Beyotime Biotech) was used for the EdU assay. Briefly, the EdU reagent was directly dissolved into the culture medium at a 10-μM final concentration. After incubation for another 2 h, the CRC cells were fixed with paraformaldehyde for 15 min and stained with Click Additive Solution in the dark. Thereafter, Hoechst staining was used to counterstain nuclei, and EdU-positive cells were photographed and counted under a fluorescence microscope.

Luciferase reporter assay

The targeted binding sites among circ_0003215, miR-663b, and DLG4 were confirmed by the dual-luciferase reporter assay. The binding sites between circ_0003215 and miR-663b, DLG4 and miR-663b were predicted by bioinformatics analysis. Luciferase vectors with circ_0003215 or the 3′-UTR of DLG4 and mutant types containing the firefly and Renilla luciferase genes were obtained from Promega (Madison, WI, USA). Next, HEK-293T cells were transfected with psiCHECK-circ_0003215 vectors (wild-type [WT] or mutant [MUT]) and miR-663b mimics, or psiCHECK-DLG4 vectors (WT1, WT2, or MUT1, MUT2) and miR-663b mimics. Relative luciferase activities were determined according to the corresponding instructions for the Dual-Luciferase Reporter Assay kit (Promega).

RNA immunoprecipitation (RIP) assay

The RIP assay was conducted using the EZ-Magna RIP kit (MilliporeSigma, Burlington, MA, USA). Briefly, CRC cells were incubated with RIP lysis buffer containing RNase and protease inhibitors on ice for 60 min. Next, the cell lysates were incubated overnight with RIP buffer containing magnetic beads conjugated to IP-grade antibodies. Subsequently, the immune complexes were washed five times and treated with proteinase K. Finally, the immune precipitated RNA was isolated and subjected to RT-qPCR analysis.

RNA pull-down assay

The biotinylated-circ_0003215 and NC probes were obtained from GenePharma (Shanghai, China). Briefly, CRC cells (1.5 × 107) were fixed and lysed in lysis buffer. Afterward, the supernatant was collected, and 50 μL was pipetted for input after centrifugation, with the remainder incubated with biotin-labeled probes. Subsequently, the complexes were incubated with streptavidin-coupled Dynabeads overnight. Next, the beads that captured the immune complex were washed twice and then treated with proteinase K at 25 °C for 1 h. Finally, the bound RNAs were extracted using TRIzol, followed by RT-qPCR analysis. All probe sequences are listed in Supplementary Table S4.

Oxidative PPP flux assay

Oxidative PPP flux was determined by measuring 14CO2 release. Briefly, a 10-cm dish with two sealed pinholes containing a 6-cm Petri dish was seeded with CRC cells. Afterward, the cells were cultured for 4 h with medium supplemented with [1-14C] or [6-14C]glucose until 80% confluence. The 14CO2 released from the cells was harvested by sealing the 10-cm dish. Oxidative PPP flux was stopped by injecting 2 N HCl through one of the holes, and the released 14CO2 was harvested for 12 h by injecting hyamine hydroxide into a cup placed on the 10-cm dish through the other hole. Subsequently, the hyamine hydroxide in the cup was dissolved in 60% methanol and subjected to scintillation counting. The 14CO2 content released was obtained by subtracting the scintillation count of the [6-14C] glucose treatment from that of the [1-14C]glucose treatment.

ROS

The amount of intracellular ROS was quantified using dichlorodihydrofluorescein diacetate (DCFH‐DA; Invitrogen) oxidation. CRC cells transfected with specified plasmids were seeded into 6-well plates. After 24 h, the cells were washed and subsequently treated with 10 μM DCFH-DA for 30 min. Finally, the cells were harvested, resuspended in 500 µL FACS buffer, and subjected to FACS analysis (BD Biosciences).

NADPH/NADP+ ratio

The NADPH/NADP+ ratio was determined using a Colorimetric Assay Kit (Sigma-Aldrich). Briefly, 3 × 106 transfected cells were harvested and lysed with 200 μL NADP+/NADPH extraction buffer. Thereafter, the lysed cells were incubated for 5 min at 60 °C, and then 20 μL assay buffer and 200 μL counter NADPH/NADP+ extraction buffer were added. Next, the samples were centrifuged for 20 min, and the supernatants were used to determine the NADPH/NADP+ ratio. Finally, the absorbance at 565 nm was determined using a plate reader at 0 and 30 min.

Protein half‑life assay

Cells were seeded in 6-well plates, and when they reached approximately 70% confluence, they were treated with the protein synthesis inhibitor cycloheximide at the indicated times. Afterward, the cells were harvested and analyzed by immunoblotting with specific antibodies.

Immunoprecipitation

For immunoprecipitation, transfected cells were lysed using RIPA lysis buffer containing protease inhibitors on ice for 2 h. Next, 900 μL samples were incubated with protein A/G magnetic beads conjugated with immunoprecipitation antibodies at 4°C overnight. Finally, the bead-antibody complexes were washed thrice with RIPA lysis buffer, and the immuno-complexes were subjected to western blotting analysis.

Ubiquitination assay

Cells were transfected with specific plasmids (HA-WT Ub, HA-K6O Ub, HA-K11O Ub, HA-K27O Ub, HA-K29O Ub, HA-K33O Ub, HA-K48O Ub, HA-K63O Ub, etc.). Forty-eight hours later, MG132 (10 μM; Sigma-Aldrich) was directly added to the culture media, and the cells were incubated for an additional 6h. Thereafter, the cells were collected, and the target protein were immunoprecipitated using A/G beads and IP-grade antibodies. Finally, the eluted proteins were analyzed by western blotting.

Xenograft nude mouse model

Animal studies were approved by the Institutional Animal Care and Use Committee. All BALB/c mice were kept at the Wuhan Institute of Virology, Chinese Academy of Sciences (Wuhan, China), and were strictly and equally randomized. For the tumorigenesis assay, CRC cells with circ_0003215 overexpression or the respective control cells were suspended in 100 µL PBS and injected into the back of each mouse. Subcutaneous tumor volumes were estimated every 3d using the formula: volume = 1/2 × length × width2. Three weeks later, all mice were sacrificed, and subcutaneous tumors were excised and weighed. For the lung metastasis model, CRC cells (5 × 105) with different transfections were injected via the caudal vein. All tail vein-injected mice were sacrificed after 6 weeks. Subsequently, their lungs were extracted and weighed, and the number of metastatic nodules was carefully calculated, experimenters were blinded from the treatment paradigm and where not involved in data analysis.

Bioinformatics analysis

The sequence data of all circRNAs were searched and downloaded from circBank. The target miRNAs that directly interacted with circ_0003215 were predicted using circBank. Target miRNA genes were predicted using Targetscan, miRWalk, and miBase.

Statistical analysis

All experiments were carried out independently at least thrice. The data are presented as mean ± standard deviation. Statistical analysis were conducted using SPSS 19.0 (SPSS Inc, Chicago, IL, USA) and GraphPad Prism v8.3.0 (GraphPad Software, San Diego, CA, USA). Before statistical analysis, the homogeneity of variance between groups was tested. The correlation between clinicopathologic parameters and circ_0003215 expression was analyzed using the chi-square test. Two-group comparisons of continuous variables were analyzed using the paired or unpaired Student’s t-test, all data followed a normal distribution. P < 0.05 was considered statistically significant.