TL;DR
This study provides the first comprehensive, transcriptome-wide map of how circular RNAs (circRNAs) interact with RNA-binding proteins (RBPs) across different cell types. They discovered that circRNAs are not just passive molecules but actively engage with RBPs, influencing their own production and impacting cancer-related gene expression, with implications for patient survival.
What Problem Does This Paper Address?
Circular RNAs (circRNAs) are a fascinating class of RNA molecules that form a closed loop, making them highly stable. While their potential to regulate gene expression and act as biomarkers is increasingly recognized, our understanding of how they function at a molecular level is still limited. Specifically, it was largely unknown how widespread and important the interactions between circRNAs and RNA-binding proteins (RBPs) are. RBPs are crucial for almost all aspects of RNA biology, and their dysregulation is linked to many diseases, including cancer. This paper aims to fill this knowledge gap by systematically investigating circRNA-RBP interactions across the entire transcriptome and exploring their functional consequences, particularly in the context of cancer.
Methods
The researchers employed a multi-pronged approach:
- Large-scale circRNA and RBP profiling: They analyzed deep RNA sequencing data from two human cell lines (HepG2 and K562) to identify and quantify circRNAs. Simultaneously, they utilized extensive eCLIP (enhanced CLIP) data, which maps RBP binding sites across the genome with high resolution, for 150 different RBPs.
- Bioinformatic analysis: They developed sophisticated computational methods to identify circRNA-RBP interactions by overlapping circRNA sequences with RBP binding sites. This included analyzing RBP binding in both exonic regions of circRNAs and in flanking intronic regions, which are crucial for circRNA biogenesis.
- Experimental validation: Promising circRNA-RBP interactions were validated using RNA immunoprecipitation (RIP) assays.
- Functional characterization: To understand the biological impact of these interactions, they performed knockdown experiments (using siRNA) of specific circRNAs and RBPs, followed by RNA sequencing to assess changes in gene expression. They also investigated the role of circCDYL in bladder cancer cell lines and analyzed clinical data to correlate circCDYL levels with patient survival.
Datasets
- ENCODE cell lines: Deep RNA sequencing data from HepG2 (n=5) and K562 (n=11) cell lines were used for circRNA profiling.
- ENCODE eCLIP data: Binding sites for 150 RBPs in HepG2 and K562 cell lines were utilized.
- Bladder cancer cell lines: Data from J82 and UMUC3 cell lines were used for functional studies.
- Bladder cancer patient cohort: Clinical data from 56 patients with non-muscle invasive bladder cancer were analyzed for survival correlations.
Key Insights and Ideas
- CircRNAs are RBP hotspots: Contrary to previous predictions, circRNAs are significantly enriched with RBP binding sites compared to their linear counterparts, even when accounting for gene expression levels. This suggests circRNAs are not just passive molecules but actively engage with the cellular machinery.
- Cell-type specificity is key: CircRNA-RBP interactions are highly cell-type specific, meaning a circRNA might interact with different RBPs in different cell lines or tissues. This highlights the complexity of RNA regulation.
- RBPs influence circRNA biogenesis: The study identified KHSRP as an RBP that binds to intronic regions flanking circRNAs and affects circRNA production, suggesting a role in the biogenesis pathway.
- circCDYL as a prime example: The circRNA circCDYL, implicated in bladder cancer, is almost entirely covered by RBP binding sites. Its interactions with RBPs like GRWD1, IGF2BP1, and IGF2BP2 appear to play a significant role in regulating cancer-related pathways and gene expression.
- CircCDYL levels predict survival: Higher levels of circCDYL in bladder cancer patients were associated with better overall survival, suggesting its potential as a prognostic biomarker.
Key Results
- A comprehensive catalog of circRNA-RBP interactions was generated for HepG2 and K562 cells.
- KHSRP binding sites were enriched in introns flanking circRNAs, and KHSRP depletion affected circRNA biogenesis.
- Exons comprising circRNAs showed significantly higher RBP coverage than non-circulating exons, with some RBPs showing preferential binding to circRNAs.
- Experimental validation confirmed numerous predicted circRNA-RBP interactions, demonstrating their biological relevance.
- Knockdown of circCDYL in bladder cancer cells led to differential expression of 2624-3014 genes, affecting key cancer pathways like MAPK signaling.
- circCDYL depletion had opposite effects on certain genes compared to GRWD1 depletion, supporting an RBP sponge mechanism.
- Elevated circCDYL levels were significantly correlated with improved overall survival in bladder cancer patients.
Limitations and Future Work
- Distinguishing linear vs. circular RNA binding: The eCLIP data, generated from total RNA, cannot definitively distinguish whether an RBP binds to the circRNA itself or its linear counterpart. While the study used various approaches to infer circRNA-specific binding, this remains a challenge.
- eCLIP read length: The relatively short read lengths in eCLIP data made it difficult to definitively map RBP binding sites directly across the backsplice junction (BSJ) of circRNAs.
- Functional complexity: The study acknowledges that circRNA-RBP interactions are complex and can involve multiple RBPs, competition for binding sites, and dynamic interactions. Further research is needed to fully elucidate these mechanisms.
- Clinical translation: While circCDYL shows promise as a biomarker, further investigation is required to understand its precise regulatory role in tumorigenesis and its full clinical potential.
- Future directions: The authors suggest further investigation into how circRNAs mechanistically control RBP localization, activity, and homeostasis, and how these interactions contribute to global health and disease. They also emphasize the need for more research into the clinical value of circRNAs as biomarkers.
Summary updated