Frontotemporal dementia strikes adults in their 50s and 60s — often younger than Alzheimer's disease — and currently has no disease-modifying treatments. A new preclinical tool could accelerate the path toward one. By embedding a human genetic fault directly into a mouse model, researchers have created a more clinically faithful testing ground for therapies targeting progranulin haploinsufficiency, one of the most common heritable causes of FTD.
The team engineered mice carrying a human GRN transgene with a specific four-base-pair deletion in exon 5 (GRNc.388_391delCAGT), a mutation known to cause FTD in humans. This truncated progranulin protein — designated GRNmEx5 — is expressed at reduced levels yet retains partial function, providing a partial rescue of neuropathological and transcriptomic abnormalities seen in progranulin-null animals. The critical innovation was pairing this humanized transgene with a mouse progranulin knockout background (Grn⁻/⁻; GRNmEx5), creating animals dependent entirely on the mutant human protein. Using CRISPR/Cas9 delivered via lipid nanoparticles, the researchers achieved 8.5% homology-directed repair of the target mutation in vivo — demonstrating both the deliverability of the editing machinery and the model's utility as a screening platform.
The 8.5% correction rate is modest but meaningful in context: even partial restoration of progranulin expression has been shown in prior work to confer functional benefit, since the disease mechanism is haploinsufficiency rather than dominant toxicity. Progranulin-directed therapies are an active area, with antisense oligonucleotides and small molecule approaches also in development, and this humanized model fills a genuine gap — existing mouse models carried only murine Grn mutations, limiting translational fidelity for human-sequence-targeting therapies. Key limitations include the inherent differences between mouse and human neuroanatomy, single-exon focus, and the fact that 8.5% editing efficiency will likely require substantial improvement before clinical consideration. Still, as a defined, reproducible pre-clinical platform carrying the exact human pathogenic sequence, this model represents a genuinely useful infrastructure advance for the FTD gene therapy field.