Template-free CRISPR-Cas restores hearing in deaf mice

Frameshift mutations account for more than 20% of Mendelian inherited diseases, yet existing gene-editing therapies lack efficient, broadly applicable correction tools. TIGER exploits the observation that SpCas9-mediated double-strand breaks induce reproducible insertion/deletion (InDel) patterns that can be predicted from the surrounding nucleotide context. By systematically identifying the factors – in-frame ratio and length-restoring ratio – that govern therapeutic efficacy across cells and tissues, the authors developed a gRNA scoring system and an online prediction tool, the TIGER portal, based on a fine-tuned inDelphi algorithm.

Approximately 75% of single-nucleotide deletion mutations and 50% of insertion mutations achieved the threshold of ≥30% in-frame products considered sufficient for phenotypic rescue; 38% and 65%, respectively, produced genuine wild-type protein. For in vivo validation, dual AAV9 vectors carrying SpCas9 and an optimised gRNA targeting the Otof c.1233delC mutation were injected into the inner ears of neonatal deaf mice. At three months post-injection, approximately 44% of inner hair cells expressed Otoferlin, auditory brainstem response thresholds recovered to roughly 30 dB SPL, and behavioural hearing tests confirmed functional restoration. Therapeutic effects were sustained over 12 months of follow-up, outperforming gene-replacement therapy, which showed expression decline and hair-cell loss from nine months onwards. Whole-genome sequencing detected no off-target editing, and TIGER editing efficiency was approximately 30-fold higher than prime editing in a parallel comparison.

The study was led by Wei Xiong at the Chinese Institute for Brain Research, Beijing. It was published on 23 March 2026 in Nature Biomedical Engineering.