Toxicological Blind Spots of Base Editing: Cell-Type Specific Chromosomal Translocation Risks of CBE Identified in an FHL3 Model
##1. Familial Hemophagocytic Lymphohistiocytosis Type 3 (FHL3) and Unc13d Splicing Mutation Familial hemophagocytic lymphohistiocytosis type 3 (FHL3) is a lethal hereditary hyperinflammatory syndrome caused by defects in the Unc13d gene that disrupt the degranulation mechanism of cytotoxic T lymphocytes (CTLs). In particular, the 'cryptic splice-site mutation' modeled in the Jinx mouse impedes normal protein translation, triggering a fatal cytokine storm. Conventional gene‑addition vector delivery faces challenges in expression control, necessitating a precise correction strategy that directly restores the endogenous splicing sequence within the genome.
##2. Genetic Rescue of Diseased Cells via Cytosine Base Editing (CBE) The investigators employed a cytosine base editor (CBE) that does not induce double‑strand DNA breaks (DSBs) to directly target the aberrant splice site of Unc13d. Across fibroblasts and clinically challenging cell types such as T cells and hematopoietic stem cells (HSCs), they achieved overwhelming genome‑editing efficiencies ranging from 62 % to 89 %. This restored normal Unc13d splicing, re‑established CTL cytotoxic function, and models receiving corrected HSC transplants were fully protected from virus‑induced hyperinflammatory responses.
##3. Reversed Safety Profile: Structural Variants and Chromosomal Translocations Driven by Hyperactive CBE Genotoxicity profiling, however, flags hidden hazards of base editing. A CBE platform with hyperactive deaminase activity induces far more extensive guide‑RNA‑dependent and -independent off‑target editing than conventional CRISPR‑Cas9, generating unexpected large‑scale structural variants. Notably, the long‑term stability of CBE‑induced chromosomal translocations varies dramatically across cell types, demonstrating that the same editing tool can produce markedly different toxic outcomes depending on chromatin architecture and DNA‑repair pathways in the target cell.
##4. Shift from Tool‑Centric to Cell‑Centric (Context‑Specific) Safety Assessment The pivotal significance of this work lies in redefining regulatory criteria for gene‑therapy clinical entry—from evaluating the intrinsic toxicity of the editing tool to assessing context‑specific genotoxicity for each cell type. The absence of DSBs alone should not engender complacency regarding base‑editor safety; independent genotoxicity and translocation screening must precede ex vivo products, particularly hematopoietic stem cells (HSCs) and T cells. This represents a critical technical inflection point for meeting future FDA guidelines on advanced regenerative medicine therapies (RMAT).
Source: Genome Editing & Advanced Therapeutics, May 2026. DOI: 10.1038/s41588-026-03145-y
Summary: This study applies cytosine base editing (CBE) to correct a cryptic splice-site mutation in the Unc13d locus, successfully restoring cytotoxic T-cell function and rescuing FHL3 hyperinflammatory mouse models with 62%-89% efficiency. However, comparative genotoxicity profiling reveals that hyperactive CBE induces broader off-target modifications and structural variants than CRISPR-Cas9. Crucially, the stability of CBE-induced chromosomal translocations exhibits distinct cell type-specific patterns, highlighting the imperative for context-dependent safety profiling in clinical translation.
These data quantify the cell‑type‑specific genotoxicity of the next‑generation gene‑editing tool CBE, establishing a robust safety benchmark for the genomics community. By demonstrating the curative potential for the severe hyperinflammatory disorder FHL3 and supplying cell‑specific chromosomal translocation pattern data, the work provides a critical training set for future AI‑driven off‑target prediction engines and for designing safety‑validation pipelines for cell‑based therapeutics.