circCAPRIN1 interacts with STAT2 to promote tumor progression and lipid synthesis via upregulating ACC1 expression in colorectal cancer

2.1 Patients and tissue samples

A total of 91 pairs of human CRC and adjacent non-tumor tissues were collected from Fudan University Shanghai Cancer Center (FUSCC, Shanghai, China) between September 2020 and April 2021 for quantitative PCR (qPCR) analysis. Tissue microarrays (TMA) were collected for in situ hybridization (ISH) analysis from patients with CRC admitted to FUSCC between January 2007 and November 2009. This TMA was constructed as previously described [54]. All samples were obtained from surgical resections for CRC patients and reviewed by two independent pathologists from FUSCC. The clinicopathological data of TMA are summarized in Table 1. Overall survival (OS) was defined as the duration from surgery date to death due to any cause or the recent follow-up. Disease-free survival (DFS) was calculated from the date of surgery to disease relapse or the recent follow-up.

2.2 Cell lines and cell culture

Human CRC cell lines LoVo, HCT116, SW480, SW620, DLD1, RKO, and HCT8 were purchased from purchased from Cell Bank of Type Culture Collection of the Chinese Academy of Sciences in Shanghai, China and cultured in DMEM medium (Cytiva, Marlborough, USA) supplemented with 10% fetal bovine serum (Gibco, Invitrogen, USA) and 1% penicillin/streptomycin (Gibco, Invitrogen, USA), while the intestinal epithelial cell line NCM460 was purchased from purchased from Cell Bank of Type Culture Collection of the Chinese Academy of Sciences in Shanghai, China and cultured in RPMI 1640 medium (Cytiva, Marlborough, USA) supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin. All cells were incubated at 37°C with 5% CO₂.

siRNAs were transfected into CRC cells using Lipofectamine 3000 reagent (Invitrogen, California, USA) according to the manufacturer's protocol. The siRNA sequences targeting STAT2 were listed in the Supplementary Table S1.

HEK293T cells were co-transfected with pLKO.1 (lncbio, Shanghai, China) or pLenti ciR6024 (lncbio, Shanghai, China) and the packaging plasmids psPAX2 (lncbio, Shanghai, China) and pMD2G (lncbio, Shanghai, China). The lentiviral particles were collected and filtered via a 0.45-μm membrane (Merck Millipore, Darmstadt, Germany) after 72 h post-transfection. The lentiviral particles and polybrene (10 μg/mL; Beyotime Biotechnology, Shanghai, China) were added to CRC cells, and infected cells were selected under puromycin (Beyotime Biotechnology, Shanghai, China) before harvest or subsequent experiments.

2.3 CCK-8 and colony formation assays

For CCK-8 assays, 1000 cells/well from different groups were seeded in 96-well plates. CCK-8 (Meilunbio, Dalian, China) reagent was added to each well at a fixed time each day for five days, and the cells were incubated at 37°C for 2 h. Then, the absorbance was measured in the single-wavelength mode (450nm) by microplate reader (Biotek, Synergy H1, Vermont, USA).

For colony formation assays, 500 cells/well were seeded in 6-well plates for each group. When visible colonies were formed, the cells were washed with phosphate-buffered saline (PBS) and stained with 0.1% crystal violet for 5 min. Subsequently, the visible colonies were counted.

2.4 Cell migration assays

The wound healing and transwell assays were used to detect the migration ability of CRC cells. In the wound healing assay, RKO and LoVo cells (5 × 10⁵) cells were cultured in six-well plates and scratched with a sterile pipette tip when each well was filled with cells. After that, the cells were cultured in serum-free DMEM medium and the wound size was measured at 0 and 24 h. In transwell assay, RKO and LoVo cells (5 × 10⁴) suspending in 200 μL serum-free DMEM medium were cultured in upper chambers, and 500 μL 10% FBS-medium was filled in the lower chamber. After 24 h incubation, 0.1% crystal violet were used to stain the cells covered in the bottom chamber. After removing the cells in the upper chamber, the residual cells adhered to the bottom chambers were photographed.

2.5 RNA extraction and real-time qPCR

Total RNAs were extracted from tissues and cells using TRIzol (Invitrogen, California, USA) and Isopropanol (Aladdin, Shanghai, China), and reverse transcription was performed using PrimeScript RT Master Mix (Takara, Osaka, Japan). Real-time qPCR was performed using SYBR Premix (EZBioscience, Roseville, USA). The primers are listed in Supplementary Table S1.

2.6 Plasmid construction

HcircCAPRIN1 was PCR-amplified and inserted into pLenti ciR6024 vector (lncbio, Shanghai, China). shRNA targeting circCAPRIN1 were generated and inserted into pLKO.1 vector (lncbio, Shanghai, China). The primers and shRNA sequences used in this study are summarized in Supplementary Table S1.

2.7 Actinomycin D and RNase R treatment

CRC cell lines were seeded in a 24-well plate and cultured overnight. Total RNAs were extracted from cells using TRIzol (Invitrogen, California, USA) and Isopropanol (Aladdin, Shanghai, China) after actinomycin D (2 mg/mL; MedChemExpress, Shanghai, China) or DMSO (Solarbio, Beijing, China) treatment for different periods (0, 12, and 24 h). An equivalent of 5 μg RNA was incubated with or without RNase R (20 U; Thermo Fisher Scientific, Waltham, USA) at 37°C for different periods (0, 10, 20, 30, and 40 min). After treatment with actinomycin D or RNase R, the relative expression of circCAPRIN1 or CAPRIN1 mRNA was detected by PCR or qPCR. The sequences of indicated primer used in qPCR are summarized in Supplementary Table S1.

2.8 Oil Red O and Nile Red staining

CRC cells grown on coverslips were fixed with 4% (wt/vol) paraformaldehyde (PFA; Sangon Biotech, Shanghai, China) for 15 min and stained with 3 mg/mL Oil Red O solution (Yuanye, Shanghai, China) for 30 min. After nuclear counterstaining with hematoxylin (Yuanye, Shanghai, China), lipid droplet was visualized under a microscopy (Olympus, DP73, Tokyo, Japan) and evaluated using Image J software (National Institutes of Health, Bethesda, USA).

Similarly, CRC cells grown on coverslips were fixed with 4% (wt/vol) PFA for 15 min and stained with 2 μmol/L Nile Red solution (Solarbio, Beijing, China) for 30 min. Subsequently, the nuclei were stained with 4,6-diamidino-2-phenylindole (DAPI; Thermo Fisher Scientific, Waltham, USA), and immunofluorescence signals were visualized under confocal microscopy (ZEISS, LSM880, Oberkochen, Germany).

2.9 Western blotting

Total proteins were extracted using Radio Immunoprecipitation Assay lysis buffer (Beyotime Biotechnology, Shanghai, China) supplemented with protease inhibitors (Selleck, Shanghai, China) and phosphatase inhibitors (Selleck, Shanghai, China). An aliquot of protein (40 μg) was separated by Sodium dodecyl-sulfate polyacrylamide gel electrophoresis (SDS-PAGE; 10%) and transferred to a PVDF membrane (Bio-Rad Laboratories, Shanghai, China). The membranes were blocked with 5% milk and probed with the primary antibodies at 4°C overnight, followed by incubation with the secondary antibodies (1/5000; #7074 or #7076; Cell Signaling Technology) at room temperature for 1 h. The primary antibodies used in this study were as follows: CPT1A (1/1000; ab220789; Abcam), STAT2 (1/500; sc-1668; Santa Cruz Biotechnology), Snail (1/1000; #3879; Cell Signaling Technology), N-Cadherin (1/1000; #14215; Cell Signaling Technology), E-Cadherin (1/1000; #14472; Cell Signaling Technology), ACC1 (1/1000; 21923-1-AP; Proteintech), ACC2 (1/1000; #8578; Cell Signaling Technology), GAPDH (1/1000; #2118; Cell Signaling Technology), p-ACC (1/1000; #11818; Cell Signaling Technology), Vimentin (1/1000; #5741; Cell Signaling Technology), and FAS (1/1000; #4233; Cell Signaling Technology)., p-STAT2 (1/1000; #ab191601; Abcam); TOFA (3.3 μg/mL; #S6690; Selleck); Cerulenin (10 μg/mL; #HY-A0210; MedChemExpress). Then, the membranes were exposed with ECL reagent (Beyotime, Shanghai, China) for chemiluminescent detection.

2.10 Fluorescence in situ hybridization (FISH) and immunofluorescence (IF) staining

FISH was conducted to detect circCAPRIN1 in CRC cells using the FISH Probe Mix (green; GenePharma, Shanghai, China). First, RKO and LoVo cells were immobilized by 4% formaldehyde for 15 min, followed by incubation with FAM-labeled probe against circCAPRIN1 at 37°C overnight. The sequence of the probe is listed in Supplementary Table S1. Second, the cells were incubated with antibodies specific for ACC1 (1/200; 21923-1-AP; Proteintech) and STAT2 (1/50; Santa Cruz Biotechnology, sc-1668) at 4°C overnight, and then with Alexa Fluor 594 AffiniPure goat anti-rabbit IgG (1:200; 111-585-003; Jackson ImmunoResearch, West Grove, PA, USA) and Alexa Fluor 647 AffiniPure goat anti-mouse IgG (1:200; 115-645-003; Jackson ImmunoResearch, West Grove, PA, USA) at 37°C for 2 h. Subsequently, the slides were washed with PBS. Finally, the nuclei were stained with DAPI (Thermo Fisher Scientific, Waltham, USA), and the IF signal was visualized under a confocal microscopy (ZEISS, LSM880, Oberkochen, Germany).

2.11 In situ hybridization (ISH)

TMA was dewaxed in xylene and rehydrated sequentially with 100%, 95%, 85%, and 75% alcohol. Then, the tissues were hybridized with a specific digoxin-labeled circCAPRIN1 probe (Servicebio, Wuhan, China). The sequence of the probe is listed in Supplementary Table S1. The expression of circCAPRIN1 was quantified and analyzed. The positive staining intensity (0, negative; 1, weak; 2, moderate; 3, strong) was multiplied by the percentage of positive staining cells (0, < 10%; 1, 10%-25%; 2, 26%-50%; 3, 51%-75%; 4, > 75%) to calculate the score for ISH staining. The ISH score ≤ 8 defined low expression of circCAPRIN1, while > 8 indicated high expression of circCAPRIN1.

2.12 Transmission electron microscopy (TEM)

For TEM, CRC cells were washed with PBS and fixed with 2.5% pentadiol (Servicebio, Wuhan, China) before analysis at 120 kV using TEM (Hitachi HT7700). Briefly, CRC cell lines were fixed with 2.5% pentadiol (Servicebio, Wuhan, China) at room temperature for about 5 min. Then the cells were detached gently with cell scrapers and were centrifuged (no more than 3000 rpm/min) for about 2 min. Finally, CRC cells were resuspended with 2.5% pentadiol and were observed under TEM.

2.13 RNA pull-down, silver staining, and mass spectrometry analysis

Biotin-labeled probes were synthesized by Shanghai GenePharma Co., Ltd (Shanghai, China). The sequence of the probe is listed in Supplementary Table S1. RNA pull-down assay was performed using the Pierce Magnetic RNA-Protein Pull-Down Kit (Thermo Fisher Scientific, USA) according to the manufacturer's protocols. circCAPRIN1-binding proteins were resolved on SDS-PAGE (10%) and stained using Fast Silver Stain Kit (P0017S; Beyotime), following the manufacturer's instructions. Subsequently, the proteins were identified by mass spectrometry (Shanghai Applied Protein Technology Co., Ltd) and verified by Western blotting analysis. Liquid chromatography-mass spectrometry analysis was performed on a Q Exactive mass spectrometer (Thermo Scientific, Waltham, USA) coupled to Easy nLC (Proxeon Biosystems, now Thermo Fisher Scientific) for 120 min. The mass spectrometer was operated in positive ion and peptide recognition modes. Mass spectrometry data was acquired using a data-dependent top 10 method dynamically choosing the most abundant precursor ions from the survey scan (300-1800 m/z) for high-energy collisional dissociation HCD fragmentation. Automatic gain control (AGC) target was set to 3e6, and maximum inject time to 10 ms. Dynamic exclusion duration was 40.0 s. Survey scans were acquired at a resolution of 70,000 at 200 m/z and resolution for HCD spectra was set to 17,500 at 200 m/z, and isolation width was 2 m/z. Normalized collision energy was 30 eV and the underfill ratio, which specifies the minimum percentage of the target value that to could be reached at the maximum fill time, was defined as 0.1%. The instrument was run with peptide recognition mode enabled.

2.14 RNA immunoprecipitation (RIP) assay

STAT2 primary antibody (2 μg; sc-1668; Santa Cruz Biotechnology) was selected for RIP assay using EZ-Magna RIP Kit (Merck Millipore, Darmstadt, Germany). In brief, cell extracts were mixed with protein A/G beads conjugated to an antibody against STAT2 or IgG (negative control). Then, the precipitated RNAs were analyzed by qPCR analysis. The sequences of the primers used are summarized in Supplementary Table S1.

2.15 Dual-luciferase reporter assay

The 2000 bp upstream of the ACC1 gene promoter was amplified using PCR and cloned into the pGL3-basic vector (Promega, Madison, USA). The sequences of the primers used are summarized in Supplementary Table S1. An equivalent of 200 ng luciferase vector pGL3-ACC1 or pGL3-mut-ACC1, 200 ng STAT2 overexpressing vector or negative control (NC) vector, and 15 ng plasmid expressing a renilla luciferase gene (pRL, Promega, Madison, USA) were transiently co-transfected into HEK293T cells and the cells were cultured in a 24-well plate. Then, the cells were lysed to detect the luciferase activity after 48 h post-transfection using Dual-Luciferase Reporter Assay System (Promega, Madison, USA) as instructions. The value of luciferase activity was detected by Synergy H1 microplate reader (Biotek, Vermont, USA). The relative firefly luciferase activity was normalized to renilla luciferase activity, and the fold changes in the reporters were calculated.

2.16 Chromatin immunoprecipitation (ChIP) assay

ChIP assay was performed using Pierce Agarose ChIP Kit (ThermoFisher Scientific, USA) to enrich 2000 bp upstream of ACC1 gene promoter region along with specific binding proteins. Primers were designed to amplify every 200 bp sequence (Supplementary Table S1) and detect the exact protein binding sites on or close STAT2 and ACC1 gene promoters. Briefly, RKO cells were fixed with 37% formaldehyde. The chromatin was fragmented by sonication on ice and immunoprecipitation was performed using antibodies against normal immunoglobulin G and STAT2 (sc-1668; Santa Cruz Biotechnology). PCR products were separated and then visualized on a 1.5% agarose gel and then visualized under an UV light.

2.17 Immunohistochemistry (IHC) and hematoxylin-eosin (HE) staining

For IHC assay, tissues were fixed with 4% PFA, embedded in paraffin, and cut into 4-μM-thick slices. The slices were baked for 24h at 58°C, deparaffinized in xylene, rehydrated in graded ethanol, quenched for endogenous peroxidase activity in 0.3% hydrogen peroxide for 15 minutes at 37°C, and subjected to antigen retrieval. Then, slices sections were blocked with 1% BSA/PBS (Sangon, Shanghai, China). Primary antibodies against ACC1 (1/400; 21923-1-AP; Proteintech), CPT1A (1/2000; ab220789; Abcam), and Ki-67 (1/600, GB121141; Servicebio) were incubated overnight at 4°C, followed by secondary antibody incubation, and incubation with horseradish peroxidase-labeled streptavidin for 15 min -. The immunocomplex was visualized with DAB, and the nucleus was counterstained with hematoxylin. For HE staining, a hematoxylin-eosin staining kit (Beyotime, Shanghai, China) was employed to directly stain the nucleus and cytoplasm of cells.

2.18 Mice and animal experiments

The protocols of animal experiments involved in this study were reviewed and approved by the Animal Ethics Committee of FUSCC (permit number: 2019FUSCCJS-092). Nude mice (4-6 weeks old) weighing 14-16 g were purchased from Shanghai Jie Si Jie Laboratory Animal Co., Ltd., (Shanghai, China) and bred under specific pathogen-free conditions. LoVo and RKO cells transfected with the above-mentioned lentiviruses were subcutaneously injected into 4-week-old nude mice to establish a mice xenograft tumor model. Each nude mouse was subcutaneously inoculated with 1 × 10⁶ cells and measured the tumor volume was measured every one week. The mice were euthanized with carbon dioxide (CO₂) 4 weeks later, then the subcutaneous tumors were harvested, weighed, and stained.

For in vivo metastasis assay, 1 × 10⁶ HCT116 cells were injected into 5-week-old nude mice via tail vein. The bioluminescent signals of liver metastasis were observed using a specific imaging system (Caliper Life Sciences, USA) after 4-5 weeks of injection. The metastatic lesions were assessed by HE staining.

2.19 circRNA microarray and RNA-sequencing (RNA-seq)

circRNA expression profiles of 3 paired CRC and adjacent cancer tissues were compared by using circRNA microarray. Total RNA from each sample was extracted using TRIzol (Invitrogen, CA, USA) following the manufacturer's instructions. The concentration and quality of total RNA were evaluated using NanoDrop ND-1000 (Thermo Fisher Scientific, Waltham,USA). Total RNA was digested with RNase R (Thermo Fisher Scientific, Waltham, USA) to remove linear RNAs and enrich circRNAs, followed by circRNAs amplification. Arraystar Human circRNA Array (8 × 15K; Arraystar, Rockville, USA) was used to hybridized the circRNAs. The arrays were then scanned by the Agilent Scanner G2505C (Agilent Technologies, California, USA). Agilent Feature Extraction software (version 11.0.1.1; Agilent Technologies, California, USA) was used to analyze the acquired array images. Quantile normalization and subsequent data processing were performed using the R software limma package.

For RNA-seq, strand-specific RNA-seq libraries were prepared using the NEBNext Ultra Directional RNA Library Prep Kit for Illumina (NEB, Beverly, MA) according to the manufacturer's instructions. The raw sequencing reads were filtered by FastQC, and compared using the spliced read aligner TopHat2. RNA-sequencing data were analyzed by spliced read aligner HISAT2 (Dusseldorf, Germany), and gene ontology (GO) analysis was carried out for gene functional annotation. Genes with a fold change of > 2 and a P value of < 0.05 was considered to be significantly differentially expressed between human colorectal cancer tissues and adjacent tumor tissues, and between RKO control and RKO shcircCAPRIN1-2/RKO shcircCAPRIN1-3 colorectal cancer cell lines.

2.20 Statistical analysis

Statistical data were analyzed using SPSS 24.0 software (IBM, Armonk, NY, USA). Categorical variables were analyzed using chi-squared test. Experimental data were analyzed using GraphPad Prism 8 software (GraphPad Software Inc., San Diego, CA, USA) and presented as mean ± standard deviation (SD). Student's t test was used to compare the statistical difference between two groups. Kaplan-Meier curves were plotted using R (version 3.6.3), and the log-rank test was employed to compare the survival differences. P < 0.05 indicated statistical significance.

3.1 circCAPRIN1 was associated with CRC

To identify the circRNAs involved in the tumorigenesis of CRC, circRNA microarray was performed using three pairs of tumor and non-tumor tissues from CRC patients. Differentially expressed circRNAs were identified, with a fold-change ≤ 0.5 and ≥ 2, and a P value < 0.05 (Figure 1A). Among these, circCAPRIN1 (hsa_circ_0021639) was most significantly upregulated in CRC tissues. CircCAPRIN1 derived from exons 8-12 of the CAPRIN1 gene on human chromosome (chr) 11 was a sense-overlapping circular transcript with 467 bp (Figure 1B). The head-to-tail splicing of the amplified circCAPRIN1 was verified by Sanger sequencing (Figure 1C). To substantiate that the loop structure of circCAPRIN1 was generated from back-splicing of pre-mRNA, we designed divergent and convergent primers for circCAPRIN1 amplification using complementary DNA (cDNA), and genomic DNA (gDNA) extracted from RKO cells as templates. As expected, circCAPRIN1 could be amplified only by using divergent primers and cDNA rather than gDNA as template (Figure 1D). FISH assay revealed that circCAPRIN1 was predominantly located in the nucleus but not cytoplasm in RKO and LoVo cells and upregulated in CRC tissues compared to normal tissues (Figure 1E). Next, the stability of circCAPRIN1 was evaluated. After treating with a transcription inhibitor actinomycin D, the half-life of circCAPRIN1 was significantly prolonged compared with linear CAPRIN1, as shown by PCR and qPCR results (Figure 1F-G). Similarly, circCAPRIN1 was more stable than linear CAPRIN1 following RNase R treatment (Figure 1H-I).

3.2 Upregulated circCAPRIN1 in CRC predicted poor prognosis

To further investigate the correlation between circCAPRIN1 expression and CRC, we first tested circCAPRIN1 expression in CRC tissues and cell lines using qPCR. The results showed that circCAPRIN1 expression was upregulated in CRC tissues compared to the paired adjacent normal tissues, which also verified our previous microarray findings (Figure 2A). Similarly, circCAPRIN1 expression in cultured CRC cell lines (LoVo, HCT116, SW480, SW620, DLD1, and RKO) was significantly higher than that in intestinal epithelial cell line (NCM460) (Figure 2B). Subsequently, the expression of circCAPRIN1 in CRC tissues was determined by ISH using human TMA (Figure 2C). The circCAPRIN1 ISH score in CRC tissues was significantly higher than that in normal tissues (Figure 2D), and was positively correlated with tumor-node-metastasis (TNM) stage and tumor grade (Figure 2E and Table 1). Patients were divided into the high- and low-expression groups according to the circCAPRIN1 ISH score (low expression: ISH score ≤ 8; high expression: ISH score > 8) (Figure 2F). Patients with high circCAPRIN1 expression had unfavorable OS (P = 0.006) and DFS (P = 0.012) (Figure 2G). These results indicated an oncogenic role of circCAPRIN1 in promoting CRC progression.

3.3 circCAPRIN1 promoted the proliferation, migration, and EMT of CRC cells in vitro

In order to investigate the biological function of circCAPRIN1 in CRC development, we used shRNAs to knock down circCAPRIN1 expression in RKO and LoVo cells. The results of qPCR analysis showed that sh-circCAPRIN1-2 (sh-circ-2) and sh-circCAPRIN1-3 (sh-circ-3) significantly downregulated the expression of circCAPRIN1 instead of CAPRIN1 (Figure 3A). FISH assay confirmed the reduction in circCAPRIN1 amount in CRC cells (Supplementary Figure S1). Subsequent Cell Counting Kit-8 (CCK-8) and colony formation assays demonstrated that decreased circCAPRIN1 expression could lead to significant inhibition of RKO and LoVo cell proliferation (Figure 3B-C). Furthermore, wound healing and transwell assays showed that the deletion of circCAPRIN1 impaired CRC cell migration (Figure 3D-E). Additionally, Western blotting analysis confirmed that downregulation of circCAPRIN1 was accompanied with increased expression of E-cadherin protein but decreased expression of N-cadherin, Vimentin, and Snail proteins (Figure 3F), indicating that circCAPRIN1 could promote EMT of CRC cells. As expected, the overexpression of circCAPRIN1 did not affect CAPRIN1 expression (Supplementary Figure S2A-B) but facilitated the proliferation ability and EMT of CRC cells (Supplementary Figure S2C-D).

3.4 circCAPRIN1 knockdown inhibited the expression of genes involved in adipogenesis

To gain an in-depth insight into the molecular function of circCAPRIN1 in CRC progression, RNA sequencing was performed to depict the gene expression profile of RKO cells following circCAPRIN1 knockdown. We then conducted GO analysis for mRNA-encoding genes to identify the potential functions of circRNAs in the occurrence and development of CRC. The enriched GO terms including biological process, cellular component, and molecular function are shown in Figure 4A. Intriguingly, adipogenesis was regulated by circCAPRIN1. According to the results of TEM and Oil Red O and Nile Red staining, downregulation of circCAPRIN1 was accompanied with decreased lipid droplets (Figure 4B-D). Overexpression of circCAPRIN1 resulted in enhanced lipid droplets, as shown by Nile Red staining (Supplementary Figure S2E). To demonstrate the relationship between circCAPRIN1-dependent lipogenesis and EMT in CRC, we treated circCAPRIN1-overexpressed RKO and LoVo cells with TOFA (3.3 μg/mL) or Cerulenin (10 μg/mL), respectively. Colony formation assay and transwell assays indicated that overexpression of circCAPRIN1 could promote proliferation and migration ability of CRC cells, and inhibition of ACC or FASN reversed this promoting effect (Supplementary Figure S3A-B). Similarly, Western blotting conformed that inhibition of ACC or FASN increased the expression of E-cadherin protein but decreased the expression of N-cadherin, Vimentin, and Snail proteins in circCAPRIN1-overexpressed cells (Supplementary Figure S3C).

Next, we searched for the lipid metabolism enzymes that mediated the role of circCAPRIN1 in adipogenesis. The data of RNA sequencing revealed that ACC1 expression was positively correlated with circCAPRIN1 level (Supplementary Figure S4). Western blotting showed that ACC1 was significantly downregulated following circCAPRIN1 depletion, while the expressions of ACC2, p-ACC, and FAS were not affected (Figure 4E), indicating that ACC1 might be a key target gene of circCAPRIN1. Moreover, the expression of CPT1A, a downstream protein of ACC1 signaling, was decreased as a result of circCAPRIN1 knockdown (Supplementary Figure S5A). To further confirm whether circCAPRIN1 promotes fatty acid synthesis via upregulating ACC1 expression, RKO and LoVo cells overexpressing ACC1 and circCAPRIN1 knockdown were established (Supplementary Figure S5B). Functional assays showed that ACC1 overexpression reversed the suppressive effects of circCAPRIN1 knockdown on CRC cell proliferation and fatty acid synthesis in these tumor cells (Supplementary Figure S6).

3.5 circCAPRIN1 interacted with STAT2 to transcriptionally activate ACC1

Since circCAPRIN1 is mainly localized in the nucleus, it might regulate transcription via interaction with specific transcription factors, thereby affecting ACC1 expression and tumor progression. Next, we conducted RNA pull-down assay, silver staining, and mass spectrometry analysis to identify circCAPRIN1-binding proteins (Figure 5A). Consistent with our hypothesis, STAT2, an interferon-regulated transcription factor, was significantly enriched in the circCAPRIN1 group instead of the lgG group (Figure 5B). FISH assay showed that circCAPRIN1 and STAT2 were colocalized in the nucleus (Figure 5C). Furthermore, RIP assay substantiated that STAT2 specifically bound to circCAPRIN1 in CRC cells, as indicated by both PCR and qPCR analyses (Figure 5D). Notably, circCAPRIN1 knockdown did not affect STAT2 expression (Figure 5E). Additionally, siRNA-mediated STAT2 knockdown resulted in decreased ACC1 level in CRC cells, suggesting that ACC1 expression was regulated by STAT2 (Figure 5F-G). Moreover, according to the GTEx database, a positive correlation between ACC1 and STAT2 levels was established in colon tissues (Supplementary Figure S7A). As expected, ACC1 expression was upregulated in colon cancer tissues compared with normal tissues according to the results from the TCGA and GTEx databases (Supplementary Figure S7B).

Next, the binding site of ACC1 was predicted, and a luciferase reporter assay was performed to confirm the correlation between STAT2 and ACC1 (Figure 5H). Compared with the control group, STAT2 overexpression significantly increased the luciferase activity of wild-type ACC1 and had a minimal effect on mutant ACC1, wherein STAT2 binding site was mutated. To identify the precise binding site of STAT2 on ACC1 gene promoter, we performed a ChIP assay in RKO cells. One STAT2 binding site was located at approximately 600-800 bp upstream of the open reading frame of ACC1 (Figure 5I). Moreover, the overexpression of STAT2 partially reversed the downregulation of ACC1 gene expression upon circCAPRIN1 knockdown and counteracted the inhibitory effect of circCAPRIN1 knockdown on the proliferation, migration, and EMT of CRC cells (Supplementary Figure S7C-E). Collectively, these data suggested that circCAPRIN1 recruits STAT2 to transcriptionally activate ACC1 gene expression in CRC cells.

3.6 circCAPRIN1 promoted CRC cell growth and metastasis in vivo

To determine the role of circCAPRIN1 in CRC in vivo, RKO and LoVo cells with stable knockdown of circCAPRIN1 were inoculated subcutaneously into nude mice. Tumor volume was measured every week. circCAPRIN1 silencing dramatically inhibited tumor growth (Figure 6A-C). IHC staining for xenograft tumors displayed weaker Ki67 staining of the tumor tissues derived from stable knockdown of circCAPRIN1 compared to the controls. Moreover, we determined the effects of circCAPRIN1 knockdown on ACC1 signaling and its downstream molecular CPT1A by IHC staining. As expected, ACC1 and CPT1A labeling was weaker after knocking circCAPRIN1 (Figure 6D).

Next, we explored the metastasis-promoting potential of circCAPRIN1 in vivo. HCT116 cell lines stably transfected with a luciferase-encoding plasmid were and injected into nude mice via tail vein. Metastatic tumors of mice were assessed using the IVIS Spectrum Imaging System. The results demonstrated that circCAPRIN1 knockdown impaired the metastatic ability of CRC cells (Figure 6E-F). Taken together, it could be deduced that circCAPRIN1 inhibited CRC cell growth and metastasis in vivo. Schematic of biogenesis, function, and regulatory mechanism of circCAPRIN1 in CRC cells is shown in Figure 7.