Base editing as a correction strategy for DMD c.2609T>G (p.Leu870Ter)
CONCLUSION
Adenine base editing (ABE) is a technically plausible approach for the DMD c.2609T>G nonsense variant. The T>G transversion is not directly addressable by current ABE or CBE windows, but cytosine base editing of the complementary strand (G>A on the sense strand reading C>T on the antisense strand) could restore the Leu codon. Careful PAM identification and window positioning are required; the alternative is a prime editing strategy targeting the precise G>T reversion. Both modalities warrant feasibility modeling before moving to cell-based validation.
EVIDENCE
DMD c.2609T>G introduces a premature stop at Leu870 within the rod domain, disrupting dystrophin mechanical function. ClinVar classifies this as Pathogenic (single submitter, criteria provided). Nonsense variants in DMD exons 10�C40 have been addressed experimentally by exon skipping and, more recently, by base editors and prime editors in patient-derived myotubes and mdx mouse models (Chemello et al. 2021, PMID:33861946; Xu et al. 2021, PMID:34380046). ABE8e and PE3 platforms have demonstrated >20% correction efficiency in cardiomyocytes derived from DMD iPSCs, supporting translational feasibility for point variants in this gene. PAM availability across exon 19�C20 junction region (where c.2609 resides) is reasonable given SpCas9-NG and SaCas9 variants with relaxed PAM requirements.
This variant lies in exon 19 of DMD, within the N-terminal rod domain (spectrin-like repeat 3). Leu870 is conserved across vertebrate dystrophin orthologs. The premature stop codon at this position produces a severely truncated protein that is rapidly degraded by NMD, resulting in a near-complete absence of dystrophin protein��consistent with the severe Duchenne phenotype rather than an intermediate or Becker presentation. Exon skipping of exon 19 alone would restore the reading frame for the subset of deletions flanking this exon, but for this point variant, exon skipping would remove in-frame coding sequence without restoring the wild-type residue, making precise correction preferable. Prime editing offers single-nucleotide resolution and does not require a double-strand break, reducing the risk of large deletions and chromosomal rearrangements reported with nuclease-based approaches at the DMD locus.
LIMITATIONS
The precise T>G transversion at c.2609 does not fall within the canonical editing window of ABE7.10 or CBE3, so prime editing or a next-generation base editor with shifted windows (e.g., ABE8e with extended window) must be confirmed by in silico modeling. Muscle delivery via AAV remains the primary bottleneck: dual-AAV split-intein strategies for large cargo (PE3 ~6.3 kb) are still under optimization, and transduction efficiency in cardiac muscle lags behind skeletal muscle. Off-target analysis by orthogonal sequencing (CIRCLE-seq, GUIDE-seq) has not been reported for this specific variant site. Long-term dystrophin restoration data beyond 6 months in animal models are lacking. No patient-specific iPSC or organoid work has been published for this allele.