TL;DR
RNA base editing tools like RESCUE offer precise genetic modifications, but their off-target effects are not fully understood. This study reveals that RESCUE induces widespread single-nucleotide variants (SNVs) across the transcriptome, impacting mRNA, circRNA, and lncRNA expression and their regulatory networks, underscoring the need for careful safety assessments.
What Problem Does This Paper Address?
RNA base editing is a powerful technology for correcting genetic mutations and advancing molecular therapies. Two prominent tools are REPAIR (A-to-I conversion) and RESCUE (A-to-I and C-to-U conversions). RESCUE's ability to perform two types of edits means it can potentially generate twice the number of mutations compared to REPAIR. However, a significant concern with base editors, particularly those performing C-to-U edits, is the induction of unintended "off-target" single-nucleotide variants (SNVs) in RNA. While some studies have investigated these off-target effects, a comprehensive understanding of their impact across the entire transcriptome (mRNA, circRNA, lncRNA, and miRNA) and their downstream consequences on gene expression and regulatory networks has been lacking. This paper aims to fill that gap by systematically characterizing the transcription-wide effects of RESCUE-induced off-target SNVs in mammalian cells.
Methods
The researchers employed a multi-pronged approach to investigate RESCUE's off-target effects:
- RESCUE Editing System: They utilized the RESCUE RNA base editor, which combines an engineered cytidine deaminase (ADAR2dd) with an inactive Cas13b ortholog (dRanCas13b).
- Experimental Design: They transfected human 293T cells with RESCUE constructs targeting either the CTNNB1 or KRAS genes, along with control groups (non-target sgRNA and eGFP).
- Transcriptome-wide Sequencing: After 48 hours, they performed high-resolution transcriptome sequencing (RNA-seq) on sorted GFP-positive cells to capture global RNA changes. This included sequencing for mRNA, circRNA, and lncRNA.
- SNV Calling and Analysis: They developed pipelines to identify and quantify both A-to-I and C-to-U off-target SNVs across the transcriptome.
- Differential Expression Analysis: They used DESeq2 to identify differentially expressed genes (DEGs), circRNAs (DECs), and lncRNAs (DELs) in the RESCUE-edited cells compared to controls.
- Correlation Analysis: They correlated the presence of off-target SNVs with changes in RNA expression.
- RNA Interaction Network Analysis: They investigated how these off-target SNVs affected the interactions between different RNA types (e.g., DEC-DEG interactions).
- Validation: Key findings were validated using RT-qPCR.
Datasets
The study utilized RNA-seq data generated from human 293T cells subjected to RESCUE-mediated RNA editing. The data encompassed mRNA, circRNA, and lncRNA expression profiles. No new benchmark datasets were introduced; rather, existing RNA-seq technologies and analysis pipelines were applied to a novel experimental system.
Key Insights and Ideas
- RESCUE is a potent generator of off-target SNVs: The study confirmed that RESCUE, particularly due to its C-to-U editing capability, induces a significant number of unintended SNVs across the transcriptome.
- Off-target SNVs are not confined to target genes: The impact of RESCUE extends far beyond the intended editing site, affecting a broad range of RNA molecules.
- Diverse RNA types are affected: RESCUE-induced off-target SNVs were found to influence the expression of mRNAs, circRNAs, and lncRNAs, demonstrating a pervasive effect on gene regulation.
- RNA interaction networks are perturbed: The study revealed that these off-target SNVs alter the complex interplay between different RNA species, suggesting downstream functional consequences.
- The need for caution and improvement: The findings highlight that while RESCUE is a powerful tool, its off-target effects necessitate careful characterization and ongoing efforts to improve its specificity for safe and effective therapeutic applications.
Key Results
- Significant SNV burden: RESCUE editing resulted in thousands of A-to-I and C-to-U off-target SNVs in the targeted cells. For instance, in the CTNNB1 editing group, 2075 A-to-I and 1424 C-to-U off-target SNVs were detected.
- Widespread gene expression changes: RESCUE-mediated editing led to the differential expression of hundreds of genes (DEGs). In the CTNNB1 editing group, 878 DEGs were identified.
- Impact on circRNAs and lncRNAs: The study identified hundreds of differentially expressed circRNAs (DECs) and lncRNAs (DELs) in RESCUE-edited cells. For example, 652 DECs were observed in the CTNNB1 editing group.
- Network alterations: RESCUE-induced off-target SNVs were associated with altered interactions between DEGs, DECs, and DELs, indicating a disruption of regulatory networks. For instance, 38 DEC-DEG interactions were found in the CTNNB1 editing group associated with A-to-I off-target SNVs.
- Location of SNVs: Off-target SNVs were found in various RNA regions, including exons, introns, and untranslated regions (UTRs).
Limitations and Future Work
The authors acknowledge that while RESCUE is a powerful tool, its off-target effects are a significant concern. They emphasize the need for careful characterization of these impacts for wider biomedical applications. The study suggests that future work should focus on developing strategies to improve the specificity of RNA editing tools. This could involve exploring approaches like conditional Cas9 expression systems or novel RNA base editor designs that minimize off-target editing, as hinted by examples like REPAIRx. Further research into the precise mechanisms by which these off-target SNVs perturb RNA networks and their functional consequences in various biological contexts is also warranted.
Summary updated