Background
Colorectal cancer (CRC), one of the most common malignancies of the digestive system, is the second leading cause of cancer-related death in the world [
1]. National Comprehensive Cancer Network (NCCN) guidelines recommend surgery combined with radiotherapy and chemotherapy as the standard treatment for CRC [
2]. Although CRC treatment has improved, the prognosis of CRC unfortunately remains unsatisfactory, primarily due to the unclear tumorigenesis mechanism and aggressive nature of CRC; approximately 880,000 CRC patients die annually [
3]. Thus, it is urgent to understand and explore the underlying mechanism of the malignant progression of CRC to improve diagnostic efficiency, treatment effect and prognosis.
Circular RNAs (circRNAs), a novel class of bifunctional RNAs that possess noncoding and limited protein-coding functions, are generated from the noncanonical back-splicing junction site of precursor mRNAs and characterized by a covalently closed single-stranded loop structure [
4]. Due to their lack of a 5′ cap and 3′ poly-A tail, circRNAs are resistant to exonucleases and more stable than their linear RNA counterparts [
5]. Moreover, with the development of high-throughput sequencing combined with available bioinformatics analysis technology, accumulating evidence has revealed that circRNAs are widely expressed, generally stable, highly conserved and tissue- or cell-specific in malignant cancers, including CRC [
6,
7]. For instance, the circRNA ciRS-7-A, circNSUN2, and circ-ERBIN were reported to be overexpressed in CRC and facilitate its progression [
8‐
10]. circRNAs can exert their biological functions by acting as microRNA (miRNA) sponges and interacting with RNA-binding proteins (RBPs). Additionally, some circRNAs can function as translation templates or scaffolds of proteins or regulators of RNA splicing and gene transcription [
11‐
13]. Aberrantly expressed circRNAs also play important roles in mediating cancer progression by regulating the activity of a variety of signaling pathways, such as the WNT/β-catenin, MAPK, JAK/STAT, NOTCH and PI3K/AKT pathways [
14,
15]. For example, Zhang et al. revealed that hsa_circ_0026628 facilitates the progression of CRC by targeting SP1 to activate the Wnt/β-catenin signaling pathway [
16]. Zheng et al. reported that circPPP1R12A promotes the pathogenesis and metastasis of colon cancer via the Hippo-YAP signaling pathway [
17]. In summary, circRNAs may serve as potential biomarkers for CRC diagnosis and prognosis prediction, as well as effective therapeutic targets and vital regulators of signaling pathways [
18].
TRIM29 (also known as ATDC), a member of the tripartite motif protein family, has been shown to play oncogenic roles in various cancers, including pancreatic cancer, thyroid cancer, bladder cancer, and CRC [
19‐
24]. Sun et al. and Han et al. demonstrated that TRIM29 is upregulated in CRC and facilitates cancer progression [
22,
23]. Xu et al. also reported that TRIM29 overexpression can promote thyroid cancer progression by activating the PI3K/AKT signaling pathway [
24]. Recently, accumulating studies have confirmed that circRNAs play a role in the tumorigenesis and progression of multiple malignancies by modulating their target genes [
9,
10]. Therefore, investigating the relationship between circRNAs and TRIM29 is of great significance for exploring the upstream regulation of TRIM29 and developing promising therapeutic targets for CRC.
In recent years, the aberrant expression and regulatory axes of circRNAs have been elucidated and described in a variety of malignancies. Nevertheless, only preliminary studies have been conducted on the role of circRNAs in CRC, and the clinical significance, levels, biological function, and molecular mechanisms of novel circRNAs in CRC remain largely unexplored. In the present study, based on bioinformatic analysis, we identified a novel CRC-related circRNA (hsa_circ_0038718) derived from exons 3 and 4 of the IL4R gene, also known as circIL4R. The expression and function of circIL4R in CRC was unclear. Hence, in situ hybridization (ISH) and qRT–PCR assays were first performed to determine the levels of circIL4R and the clinical significance of its expression. Subsequently, gain- and loss-of-function experiments were conducted in vitro and in vivo to identify the biological function and molecular mechanisms of circIL4R in CRC. More precisely, a series of experiments, such as chromatin immunoprecipitation (ChIP), luciferase reporter and biotinylated RNA pull-down assays, revealed that TFAP2C-induced circIL4R can act as a sponge for miR-761 to enhance TRIM29 expression, and forced overexpression of TRIM29 promotes ubiquitin-mediated degradation of PHLPP1 to activate the PI3K/AKT signaling pathway and promote the proliferation and metastasis of CRC cells. Collectively, our findings demonstrated that circIL4R may serve as a novel diagnostic and prognostic biomarker as well as a promising therapeutic target for CRC.
Discussion
In the past few decades, although there have been advances in the clinical treatment of CRC, its prognosis remains unsatisfactory. Therefore, identifying more reliable therapeutic targets and clarifying their effect on CRC progression are urgently needed. Recently, an increasing number of studies have investigated the biogenesis, expression, characteristics, function, and clinical significance of circRNAs in different cancers and emphasized their vital roles in regulating cancer progression [
42]. Thanks to the widespread use and availability of high-throughput sequencing and circRNA-specific bioinformatics algorithms, identification and expression data for circRNAs have been uploaded to publicly available databases, which provides opportunities for bioinformatics discovery and clinical validation. In the current study, we identified and verified a novel circRNA termed circIL4R, which was frequently upregulated in microarray analyses, CRC cell lines, and clinical samples and closely associated with malignant clinicopathological factors and an unfavorable prognosis of CRC. The detection of noninvasive biomarkers in fluids is low-cost, can be repeated and has exhibited significant clinical significance in tumor patient management [
43]. Due to their high abundance, tissue-specific expression and unique covalent circular bonded structure, circRNAs are attractive candidates for cancer liquid biopsy [
44]. In our findings, circIL4R expression was significantly different in the serum of CRC patients before and after surgical resection, as well as of healthy subjects and CRC patients. Notably, the serum expression of circIL4R exhibited a high sensitivity and specificity for CRC diagnosis and surgical efficacy monitoring, which indicated that circIL4R could serve as a liquid diagnostic biomarker. However, we must acknowledge the difficulties and challenges in introducing circRNAs to standard clinical applications. For instance, due to the presence of a back-spliced junction (BSJ) site, specific primers and probes for circRNAs are limited; thus, more sensitive and specific methods are required to increase the accuracy in the detecting of circRNAs [
6]. In addition, the identification and detection of circRNA biomarkers in most studies are only preliminary explorations in a relatively small size of sample size, and lagger clinical cohorts and prospective cohorts are required to validate the sensitivity and specificity of circRNAs in diagnosing specific diseases and predicting the prognosis of patients [
45]. Thus, optimization of the detection method, validation of the clinical significance and determination of cutoff values would be helpful for incorporating circRNAs as liquid biopsy biomarkers. Additionally, the combined detection of circRNAs and traditional markers such as CEA may be considered an effective approach. Collectively, although many obstacles need to be overcome in the future, the potential of circIL4R as a biomarker for large-scale population screening of CRC and real-time monitoring of CRC progression and treatment effects bring us promising options for liquid biopsy.
Recent studies consistently report that circRNAs can be upregulated in cancers and are correlated with tumorigenesis. Nevertheless, the upstream mechanisms of circRNA overexpression in cancers have rarely been explored. Meng et al. reported that Twist can enhance the expression of circ10720 by binding to the promoter of its host gene [
27]. Herein, our data revealed that TFAP2C mediates circIL4R overexpression via transcriptional regulation of its host gene. Our findings are consistent with previous studies reporting that TFAP2C can regulate the expression of lncRNAs by transcriptional activation [
46,
47]. Alternatively, previous studies have stated that some RBPs, such as EIF4A3 and FUS, can enhance the generation of circRNAs by binding to their flanking intron sequences [
48,
49]. Thus, whether any RBPs participate in the formation, splicing, transport or degradation of circIL4R needs to be further explored.
Our clinical analysis demonstrated that circIL4R overexpression was closely associated with tumor diameter, lymph node metastasis and distant metastasis in CRC patients, suggesting that circIL4R contributes to the proliferation and metastasis of CRC. Subsequently, in vitro gain- and loss-of-function experiments emphasized the significant roles of circIL4R in facilitating CRC cell proliferation, migration and invasion. More importantly, our subcutaneous injection model showed that circIL4R knockdown significantly reduced the tumor burden in vivo, while tail vein injection model revealed that the penetrance of metastasis to distal organs in vivo was dramatically reduced after circIL4R knockdown, including reductions in both the number and size of metastatic nodules. These in vivo and in vitro results revealed that shRNAs or siRNAs specifically targeting circIL4R may effectively inhibit CRC growth and metastasis. In general, these findings demonstrated that circIL4R may serve as an appealing therapeutic target for CRC via RNAi-based strategies. Therefore, the mechanism investigation of circIL4R driving CRC progression was urgently needed. Next, further bioinformatics analysis and validation experiments revealed that circIL4R facilitated the malignant progression of CRC via activation of the PI3K/AKT signaling pathway, which has been reported to participate in a wide range of biological processes, including the regulation of CRC progression [
31]. Xu et al. found that lncRNA MALAT1 activates the PI3K/AKT signaling pathway via the miR-26a/26b/FUT4 axis and promotes proliferation and metastasis in CRC [
50]. Kumar S et al. demonstrated that IDO1 promotes CRC cell proliferation and inhibits apoptosis via activation of the PI3K/AKT signaling pathway [
51]. Nevertheless, there is little evidence about the relationship between circRNAs and the PI3K/AKT signaling pathway in CRC.
Next, we focused on the mechanism by which circIL4R activates the PI3K/AKT signaling pathway and promotes CRC proliferation and metastasis. Since the intracellular localization of circRNA is critical to its regulatory mechanism and function, our FISH and nuclear-cytoplasmic fractionation results revealed that circIL4R is mainly distributed in the cytoplasm. Accumulating evidence shows that circRNAs located in the cytoplasm can regulate biological functions by effectively sponging miRNAs to regulate the expression of their target genes [
52]. Our previous study indicated that circSPARC overexpression upregulates the JAK2 expression by sponging miR-485-3p, thereby activating the JAK2/STAT3 pathway and enhancing CRC progression [
53]. In the current study, we observed that circIL4R can act as a sponge to competitively bind with miR-761, as verified by RNA pull-down, dual-luciferase reporter, and FISH assays. miR-761 was previously reported to be a tumor suppressor that was downregulated in several cancers and to deactivate the PI3K/AKT signaling pathway [
33,
34,
54,
55]. In agreement with previous reports, our data revealed that miR-761 expression could inhibit the proliferation and metastasis of CRC cells and was downregulated in CRC tissues and cells. Subsequent rescue experiments further revealed that miR-761 reversed the oncogenic roles of circIL4R and its ability to activate the PI3K/AKT signaling pathway in CRC. In addition to functioning as competitive endogenous RNAs, circRNAs can encode cancer-related peptides or bind with specific protein(s) to implement their functions [
11,
56]. To explore whether circIL4R could activate the PI3K/AKT signaling pathway and facilitate CRC progression by encoding protein/peptides or binding with specific protein(s), we first searched the circRNADb online database and found no open reading frame (ORF) in the sequence of circIL4R, indicating that no protein features were predicted and that the possibility of encoding protein by circIL4R was relatively low. Additionally, the cat RAPID online database predicted that circIL4R has the potential to bind with several proteins, such as SRSF9 and PTBP1. Previous studies have reported that these two proteins were involved in alternative splicing [
57,
58]. Whether circIL4R could regulate the alternative splicing via binding with SRSF9 or PTBP1 need further investigation.
Next, upon further investigation of the underlying downstream mechanism of miR-761, TRIM29 was identified as the target of miR-761 based on the cross-analysis of four microRNA targeted gene prediction databases and the publicly available TCGA and GEO databases. Notably, the interaction between miR-761 and TRIM29 has been described in previous studies. Guo et al. indicated that miR-761 facilitated the malignant phenotypes of triple-negative breast cancer cells by targeting TRIM29 [
59]. Herein, we confirmed the interaction between miR-761 and TRIM29 in CRC by qRT–PCR, western blotting and dual-luciferase reporter assays. More importantly, the results of rescue experiments further demonstrated that TRIM29 overexpression reversed the inhibition of functional phenotypes and deactivation of the PI3K/AKT signaling pathway caused by circIL4R knockdown, suggesting that the circIL4R/miR-761/TRIM29 axis facilitated the proliferation and metastasis of CRC cells by regulating the PI3K/AKT signaling pathway.
Nevertheless, to date, the mechanism by which the circIL4R /miR-761/TRIM29 axis contributes to PI3K/AKT signaling pathway activation remains elusive. Therefore, further investigations to elucidate the mechanism by which TRIM29 activates the PI3K/AKT signaling pathway are necessary. As a member of the tripartite motif protein family, TRIM29 has been shown to play oncogenic roles in various cancers, including CRC, and activate the PI3K/AKT signaling pathway in thyroid cancer [
22‐
24]. To date, whether and how forced overexpression of TRIM29 contributes to the activation of the PI3K/AKT signaling pathway in CRC remains unclear. However, the tumor suppressor genes PTEN, PHLPP1, and PP2A all inhibit the activity of the PI3K/AKT pathway [
60]. Zhou et al. reported that TRIM29 overexpression can inhibit PTEN expression and activate the PI3K/AKT pathway in nasopharyngeal carcinoma [
35]. Phillip L et al. found that TRIM29 could also induce the epigenetic silencing of PTEN in bladder cancer [
20]. Therefore, we speculated that TRIM29 may activate the PI3K/AKT pathway by regulating PTEN in CRC. Contrary to our expectations, there were no significant changes in PTEN expression when TRIM29 the expression was altered. We speculated that the reason for this phenomenon may be the specificity of cancer cells types. Intriguingly, the protein level of PHLPP1 was significantly upregulated upon TRIM29 knockdown, while the mRNA level was not changed. Moreover, the results of the rescue experiments further revealed that the deactivation of the PI3K/AKT pathway caused by TRIM29 knockdown could be reversed upon PHLPP1 knockdown. These findings suggested that TRIM29 activates the PI3K/AKT signaling pathway via posttranscriptional modification of PHLPP1. A previous study reported that the E3 ubiquitin ligase TRIM11, also a member of the tripartite motif protein family, could degrade PHLPP1 via ubiquitination to activate the PI3K/AKT pathway [
41]. We therefore examined whether TRIM29 interacted with PHLPP1. Subsequently, a series of experiments revealed that TRIM29 knockdown was accompanied by reduced levels of ubiquitinated PHLPP1, while TRIM29 overexpression was accompanied by enhanced levels of ubiquitinated PHLPP1. Therefore, we surmised that TRIM29 targets PHLPP1 for ubiquitin-mediated degradation, thereby resulting in PI3K/AKT signaling pathway activation. However, future studies are needed to fully investigate the domain required for PHLPP degradation.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.