Circular RNAs: a new frontier for cancer diagnosis and therapy

In the last decade, the extensive application of new innovative technologies has promoted the fast development of many aspects of the RNA field [1]. In particular, next-generation sequencing has helped to realize the discovery of circular RNA (circRNAs) and bring circRNAs to the center stage in cancer genomics research [2, 3]. As a new type of RNA, circRNAs are expressed endogenously as covalently closed, single-stranded circular molecules [4]. Most circRNAs are the result of a “back-splice” reaction, during which an upstream acceptor splicing site is joined with a donor splicing site [5, 6]. Originating as exons, introns, or both, circRNAs fall into three categories [7]. Exonic circRNAs (ecRNAs) originate in the sequence of pre-mRNAs during the process of exon skipping. Reverse-oriented repeats that usually appear in the introns on both sides of the exons in ecRNAs can be important for the circularization of exons [8]. Intronic circRNAs (ciRNAs) that consist of one or more introns are derived from linear RNA. Specific 11-nt C-rich element near the branch-point site and 7-nt GU-rich motifs near the 5′ splice site are vital for the biosynthesis of ciRNA [9]. Exon-intron circRNAs (ElciRNAs) are another type of circRNA stemming from both introns and exons. The circularization of EIciRNAs can be improved by the flanking repeat sequence [10]. Among the three types of circRNAs, two are located in the nucleus (ciRNAs and EIciRNA) and one is cytoplasmic (ecRNA) [11]. Most genes with circular isoforms produce only one or two distinct circRNAs, although some produce tens of distinct circular products [12]. Although the mechanism of circRNA biogenesis cannot be understood completely, the complementary sequence in the flanking introns is the basis of the biogenesis for the circRNAs of many mammals [5, 13]. Additionally, some activating or antagonistic trans-acting factors, such as quaking and ADAR, can modulate the expression of mammalian circRNAs [14, 15]. Many scientific studies have demonstrated numerous potential functions of circRNAs that clarify how circRNA changes from negative matter to its current popular rank [16]. Most circRNAs are abundant and conserved across species and exhibit tissue-specific property [17]. Importantly, circRNAs are stably expressed in the saliva, blood, and exosomes, which makes them promising biomarkers for diagnosis, prognosis, and therapeutic response for cancer patients [18]. Furthermore, in recent years, several studies have indicated that circRNAs exhibited a powerful functional potential in regulating proliferation, apoptosis, angiogenesis, and metastasis, suggesting that circRNAs may be important target molecules in cancer [19, 20]. This review briefly introduces what is currently known about the potential of circRNA as a biomarker and therapeutic target in cancer and discusses their application prospects in the clinical field.

CircRNAs are not sensitive to ribonucleases, such as exonuclease or RNase R, and have a longer half-life with a more prominent specificity in histology. Meanwhile, compared with detecting proteins through antigen-antibody reactions, testing circRNAs through quantitative real-time PCR (qRT-PCR) and in situ hybridization are more specific and sensitive. Herein, circRNAs are potential biomarkers in cancer [33].

Extracellular vesicles (EVs) are small lipid bilayer vesicles released by cells [72], which can be retooled as delivery vectors for therapeutic intervention [73, 74]. Recently, circRNAs were found to be both enriched and stable in the EVs of cancer cells [35, 75, 76]. In addition, circRNAs in cancer cells can be delivered to exosomes [77]. Furthermore, cells may use EVs to transport circRNAs to communicate to other cells [4]. Therefore, it is possible that EVs, including exosomes and microvesicles, could be engineered to deliver circRNAs efficiently to a target tissue [48, 75].

Nanoparticles have emerged as another promising approach for the delivery of circRNAs. Wu et al. has shown that, in vivo, the circular RNA circ-Foxo3 can be delivered to inhibit tumor progression by conjugating the expression plasmids of circ-Foxo3 with gold nanoparticles [78]. Since nanoparticles are not able to enter the nucleus, the therapeutic approach can only focus on exonic circRNAs, which are mainly detected in the cytoplasm.

The powerful functions and unique properties of circRNAs have made them the focus of scientific and clinical research. In this review, we summarize the recent findings on circRNAs and highlight their clinical value in cancer. First, circRNAs could act as promising biomarkers in cancer diagnosis, prognosis, recurrence, risk evaluation, and response to chemoradiotherapy. Second, circRNAs are potential therapeutic targets and promising vehicles for the delivery of therapeutics. Engineered circRNAs might be effective for sequestering oncogenic miRNAs and RBPs. Third, extracellular vesicles and nanoparticles have emerged as a promising approach for the delivery of circRNAs. In conclusion, circRNAs provide a new perspective for the diagnosis and treatment of cancer.

CircRNAs provide new insights into the “dark matter” of the human genome [79]. However, the field of circRNAs remains largely unexplored and much progress has to be made to incorporate circRNAs into clinical practice.

First, more controlled clinical studies comprising a large number of patients are required before cancer-specific circRNAs can be recommended for human use.

Second, more detailed studies on how to deliver circRNAs efficiently to the accurate action site with a long-term sustained effect and without immunological rejection are needed.

Third, the precise mechanisms of circRNA underlying the initiation and progression of cancer are worth an in-depth and extensive study, which is of significance for individual therapeutic regimens. Here are some recommendations for future research of circRNAs in cancer.

For the discovery of novel circRNA species involved in cancer, numerous algorithms have been used to detect genome-wide circRNA expression from RNA-seq data in the past few years, but there is little overlap in their predictions [12]. Herein, it is advisable to run more than one circRNA prediction algorithm to minimize the number of false positives in the absence of a gold standard algorithm.

For the stringent circRNA validation method, northern blotting is a more accurate way to fully characterize RNA species than qRT-PCR as no reverse transcription and amplification steps are part of the protocol [80]. Fluorescence in situ hybridization (FISH) provides spatial information about specific circRNAs, which is likely to be employed more in future circRNA cancer research [81].

To discover the detail binding sites for circRNA-miRNA and circRNA-RBP interactions, AGO2 RNA immunoprecipitation (RIP) combined with luciferase reporter assay, RIP followed by sequencing, and RNA pull down combined with mass spectrometry are suggested methods [82, 83].

To elucidate circRNA cellular roles, it is indispensable that exogenous synthetic molecules with engineered sequences should be designed to mimic or antagonize circRNAs. Furthermore, these molecules should be able to decoy or release miRNAs and RBPs. This research method is important to obtain circRNAs’ accurate epigenetic roles and corresponding mechanisms in cancer.

Dynamic analysis of aberrantly expressed circRNAs in cancer, such as in different stages or different sensitivities to a specific chemotherapy treatment, is very crucial for further understanding and clarifying the mechanism underlying cancer development and presenting potential therapeutic targets. A preliminary study found a large number of differentially expressed circRNAs between primary ovarian tumor and metastases using RNA-seq, which could offer a more robust classifier, for tumor subtypes or provide biomarkers for tumor screening and prognosis [84]. Another preliminary study on chemoresistant breast cancer provided several appealing targets for further functional analysis according to circRNA microarray data [47]. More comprehensive studies on those differentially expressed circRNAs will certainly gain more insights of the expression dynamics and functions of circRNAs in cancer.

We hope that one day, the appropriate and precise use of circRNAs in clinical diagnosis and treatment will become a new foothold for translational and precision medicine.