Circular RNA Sequencing: New Horizons in Biology

Once thought to be rare transcriptional byproducts, circular RNAs (circRNAs) are now recognized as important regulatory molecules with diverse biological functions. Their closed-loop structure makes them resistant to degradation, allowing them to influence gene expression and disease pathways.

Advancements in sequencing technologies have significantly improved the ability to detect and characterize circRNAs, revealing their complexity and relevance in cellular processes.

Biological Formation And Diversity

Circular RNAs arise through back-splicing, a non-canonical splicing process where a downstream splice donor covalently joins with an upstream splice acceptor. This contrasts with linear RNA splicing, which follows a sequential exon arrangement. The efficiency of back-splicing is influenced by complementary sequences in flanking introns, such as Alu repeats in primates, which facilitate pre-mRNA looping. RNA-binding proteins (RBPs) also contribute by stabilizing these interactions or modulating spliceosome activity. The interplay between these factors determines circRNA abundance and diversity across cell types and developmental stages.

CircRNAs exhibit structural diversity. Exonic circRNAs (ecircRNAs), composed entirely of exonic sequences, are primarily cytoplasmic and may regulate gene expression. Intronic circRNAs (ciRNAs), which retain intronic sequences, localize to the nucleus and may influence transcription. Exon-intron circRNAs (EIciRNAs) contain both exonic and intronic regions and have been linked to transcriptional regulation. These structural variations contribute to distinct cellular distributions and molecular interactions.

Tissue-specific expression patterns highlight the biological significance of circRNAs. High-throughput sequencing has shown that certain circRNAs are enriched in the brain, where they accumulate in synapses and contribute to neuronal plasticity. CDR1as, for example, contains multiple binding sites for microRNA-7 (miR-7) and acts as a competing endogenous RNA (ceRNA) to regulate miR-7 availability. Other circRNAs are predominant in muscle tissues, influencing myogenesis by interacting with transcription factors or modulating alternative splicing. These findings suggest circRNAs play key roles in tissue-specific regulatory networks.

Approaches To Profiling

Detecting and characterizing circRNAs requires specialized sequencing strategies that account for their back-splicing mechanism and exonuclease resistance. Advances in high-throughput sequencing have improved the ability to distinguish circRNAs from linear transcripts, enhancing understanding of their abundance, structure, and function. Various sequencing approaches, including long-read and short-read technologies, offer complementary advantages.

Long Read Methods

Long-read sequencing technologies, such as Oxford Nanopore and PacBio’s Single-Molecule Real-Time (SMRT) sequencing, enable full-length circRNA detection without extensive computational reconstruction. These platforms generate reads spanning entire transcripts, allowing direct identification of back-splice junctions and isoform diversity. Long-read sequencing is particularly useful for resolving complex circRNA structures, including those with multiple exons or retained intronic sequences, which are often fragmented in short-read datasets.

Oxford Nanopore sequencing has been widely used to characterize circRNAs due to its real-time sequencing capability and ability to detect RNA modifications. A 2022 study in Genome Biology demonstrated that Nanopore sequencing accurately distinguishes circRNAs from linear counterparts while identifying RNA modifications such as N6-methyladenosine (m6A), which may influence circRNA stability. However, long-read methods have higher error rates, requiring additional error-correction algorithms.

Short Read Methods

Short-read sequencing, primarily using Illumina platforms, remains the most widely used approach due to its high throughput and low error rate. This method involves RNA sequencing (RNA-seq) followed by computational pipelines designed to detect back-splice junctions, definitive markers of circRNA presence. Enrichment strategies, such as RNase R treatment, degrade linear RNAs while preserving circular transcripts, enhancing detection sensitivity.

Bioinformatics tools such as CIRI2, find_circ, and CircExplorer2 analyze short-read sequencing data to identify circRNAs. A 2021 Nucleic Acids Research study compared multiple detection algorithms and found that integrating multiple tools improved reliability. However, short-read sequencing struggles to reconstruct full-length circRNA sequences, especially those with multiple exons or complex splicing patterns. Computational inference of back-splice junctions can also introduce false positives, necessitating experimental validation.

Combination Strategies

Integrating long-read and short-read sequencing provides a comprehensive strategy for circRNA profiling, leveraging the strengths of both methods. Short-read sequencing offers high sensitivity for detecting back-splice junctions, while long-read sequencing enables full-length circRNA characterization and isoform resolution.

A 2023 Nature Communications study combined Illumina short-read sequencing with Oxford Nanopore long-read sequencing to profile circRNAs in human tissues. This approach identified previously unannotated circRNA isoforms and revealed alternative splicing patterns. Additionally, it facilitated the detection of RNA modifications and structural variations that would be challenging to capture using a single method. Despite these advantages, the cost and computational demands of dual-platform sequencing require optimized workflows to balance accuracy and efficiency.

Role In Gene Regulation

CircRNAs influence gene regulation at multiple levels, interacting with microRNAs (miRNAs), RNA-binding proteins (RBPs), and transcriptional machinery. Their covalently closed structure prevents rapid degradation by exonucleases, allowing them to persist and modulate gene expression over extended periods.

A key regulatory mechanism is miRNA sequestration, or sponging. Some circRNAs contain multiple binding sites for specific miRNAs, reducing their availability to suppress target messenger RNAs (mRNAs). CDR1as, for instance, harbors over 70 binding sites for miR-7, a microRNA involved in neuronal development. By binding miR-7, CDR1as prevents it from downregulating target genes, influencing synaptic plasticity and neurogenesis.

CircRNAs also serve as scaffolds for RBPs, affecting protein localization and activity. Some interact with splicing factors, influencing alternative splicing decisions. Others bind to translation initiation factors, affecting protein synthesis. Circ-FBXW7, for example, contains an internal ribosome entry site (IRES) and facilitates its own translation into a functional protein, challenging the traditional view of circRNAs as non-coding molecules.

In the nucleus, exon-intron circRNAs (EIciRNAs) and intronic circRNAs (ciRNAs) interact with RNA polymerase II and chromatin-modifying complexes to enhance transcription. EIciRNAs such as circEIF3J and circPAIP2 associate with U1 small nuclear ribonucleoproteins (snRNPs) to activate transcription of their parental genes, creating a feedback loop that influences their own expression.

Associations With Human Conditions

CircRNAs are increasingly linked to human diseases due to their stability, abundance, and regulatory functions. Their altered expression patterns suggest potential as biomarkers and therapeutic targets.

In cancer, circRNAs influence tumor progression. CircHIPK3 promotes colorectal cancer cell proliferation by sponging tumor-suppressive microRNAs, enhancing oncogenic signaling. Conversely, circSMARCA5 acts as a tumor suppressor in glioblastoma, modulating angiogenesis by interacting with RBPs involved in vascular development. These contrasting roles highlight the complexity of circRNA function in cancer biology.

Neurological disorders also exhibit distinct circRNA expression profiles. In Alzheimer’s disease, dysregulated circRNAs have been detected in brain tissue and cerebrospinal fluid, suggesting their potential as non-invasive biomarkers. A 2021 Brain study identified circRNAs correlating with amyloid-beta accumulation, a hallmark of neurodegeneration. In Parkinson’s disease, circRNAs such as circSLC8A1 influence dopaminergic neuron survival by modulating oxidative stress-related pathways. These findings suggest circRNAs contribute to neurodegeneration by affecting gene regulatory networks involved in synaptic function and neuronal maintenance.

Laboratory Validation Techniques

High-throughput sequencing provides an overview of circRNA expression, but experimental validation is necessary to confirm their presence, structure, and function. Bioinformatics pipelines can introduce false positives, making laboratory techniques essential for distinguishing genuine circRNAs.

Reverse transcription PCR (RT-PCR) with divergent primers is widely used for validation. Unlike linear RNAs, which can be amplified with convergent primers, circRNAs require primers that anneal to sequences flanking the back-splice junction in opposite orientations, ensuring specificity. RNase R digestion further confirms circRNA identity by selectively degrading linear RNAs while preserving circular molecules. Northern blotting provides additional verification by visualizing circRNAs based on their distinct migration patterns in gel electrophoresis.

Functional validation techniques help clarify circRNA interactions with proteins, microRNAs, and other cellular components. RNA immunoprecipitation (RIP) and crosslinking immunoprecipitation (CLIP) identify RBPs that associate with circRNAs, revealing their role in post-transcriptional regulation. Luciferase reporter assays assess miRNA sponge activity by measuring changes in reporter gene expression. CRISPR-Cas-based approaches have been adapted to selectively degrade circRNAs while sparing linear transcripts from the same gene, offering a precise tool for functional studies. These experimental strategies ensure that computational predictions translate into biologically meaningful discoveries.