Researchers Reveal circRNAs Act as "Molecular Sponges" Sequestering mRNA into Membraneless Granules to Regulate Gene Expression

Circular RNAs (circRNAs) are primarily generated through the back-splicing of precursor mRNAs, yet their functional targets and underlying mechanisms have remained largely elusive. Recently, researchers from the Institute of Biophysics of the Chinese Academy of Sciences published a research paper titled "Global mapping of circRNA-target RNA interactions reveal P-body-mediated translational repression" online in the journal Molecular Cell. This study introduces circTargetMap—a computational framework for the genome-wide mapping of circRNA targets using RNA-RNA interactomes obtained via RNA in situ conformation sequencing (RIC-seq) across the hippocampus and ten human cell lines. This approach identified 117,163 high-confidence circRNA-target RNA interactions, revealing that 83% of target mRNAs are bound by multiple circRNAs.

Functional investigations demonstrated that CDR1as and circRMST sequester target mRNAs into membraneless granules—specifically processing bodies (P-bodies)—through sequence-specific base pairing, thereby inhibiting the translation of these target mRNAs. This process appears to be independent of AGO2, DICER, and microRNAs (miRNAs). To directly capture these granule-associated interactions, the authors developed a method called Granule RIC-seq (GRIC-seq), which revealed a widespread role for circRNA-target RNA interactions in translational repression. Furthermore, pathogenic variants were found to be significantly enriched near circRNA-target RNA interaction sites, suggesting a potential role for these interactions in disease. This research provides a valuable resource for exploring circRNA functions and establishes an analytical framework for studying RNA-RNA interactions within membraneless organelles.

CircRNAs are a class of covalently closed RNA molecules formed by the back-splicing of precursor mRNAs. This process connects a downstream splice donor to an upstream splice acceptor site, creating a characteristic back-splice junction (BSJ). This mechanism can generate thousands of distinct circRNAs from various genes. Among them, a subset of circRNAs with high circular-to-linear expression ratios—such as CDR1as (also known as ciRS-7), circRMST, and circHIPK3—exhibit extreme stability and high abundance, and are consistently detected across different cell types and tissues, suggesting evolutionarily conserved regulatory roles. The expression of circRNAs also displays cell-type-specific, tissue-specific, and developmental stage-specific patterns, with particularly high abundance observed in the brain.

Although circRNAs have been proven to participate in critical biological processes such as differentiation, cancer, and immune regulation, their target landscapes and mechanisms of action remain unclear. Existing models propose several potential functions for circRNAs, including acting as "molecular sponges" for miRNAs and sequestering RNA-binding proteins (RBPs). The most classic example, CDR1as, contains over 70 conserved binding sites for miR-7, thereby regulating miR-7 availability. Additionally, circRNAs appear to be involved in the regulation of transcription and splicing. However, these functions are primarily derived from isolated case studies, and it remains uncertain whether a broader, generalizable mechanism exists.

Given their single-stranded structure and stability, circRNAs may also function through direct base pairing with target RNAs. Recent studies have explored this possibility using 4′-aminomethyl-4,5',8-trimethylpsoralen (AMT)-mediated psoralen crosslinking to capture circRNA-mRNA duplexes, followed by oligonucleotide pull-down and high-throughput RNA sequencing. These methods identified hundreds of circRNAs interacting with mRNAs and several circZNF609-target RNA pairs, but they are limited by low resolution, a lack of precise binding site information, and a reliance on labor-intensive pairwise validation. While bioinformatics predictions for circRNA-mRNA binding sites have been attempted, there is still a lack of a scalable, high-resolution method for the systematic mapping of circRNA-target interactions.

Figure 1. Mechanistic Diagram of circTargetMap. (LI, Peng, et al., 2026)

This study presents circTargetMap, a computational framework that globally maps circRNA-target RNA interactions by analyzing RIC-seq data—a technology capable of resolving native RNA-RNA interactomes mediated by various RBPs. By applying this framework to data from ten cell lines and human/mouse hippocampi, the authors identified 117,163 high-confidence interactions. They discovered that CDR1as and circRMST inhibit the translation of their targets through direct base pairing—a process independent of Argonaute 2 (AGO2), DICER, or miRNAs—by sequestering targets into membraneless granules such as P-bodies. To map these interactions within granules, the authors developed GRIC-seq, enabling the transcriptome-wide detection of circRNA-mRNA interactions within P-bodies. This revealed a widespread, P-body-mediated mechanism of circRNA-dependent translational repression. Moreover, the significant enrichment of pathogenic variants around circRNA-target RNA junction regions suggests that the disruption of these RNA-RNA interactions may be linked to disease.