A humanized knock-in mouse as a preclinical model for exon-skipping of a pathogenic pseudoexon causing collagen VI-related muscular dystrophy

Pre-Clinical Research
2
Veronique Bolduc Ph.D., Fady Guirguis , Sara Aguti PhD, Haiyan Zhou MD, PhD, Jun Cheng PhD, Lisa Garrett M.S., Francesco Muntoni MD, Carsten Bönnemann MD
1. NINDS/NIH, Bethesda, MD, 2. NINDS/NIH, Bethesda, MD, 3. Dubowitz Neuromuscular Centre, UCL Institute of Child Health, London, UK, 4. Dubowitz Neuromuscular Centre, UCL Institute of Child Health, London, UK, 5. Embryonic Stem Cell and Transgenic Mouse Core, NHGRI, Bethesda, MD, 6. Embryonic Stem Cell and Transgenic Mouse Core, NHGRI, Bethesda, MD, 7. Dubowitz Neuromuscular Centre, UCL Institute of Child Health, London, UK, 8. NINDS/NIH, Bethesda, MD

Collagen VI-related dystrophies are a group of frequently severe, congenital muscular dystrophies for which there currently is no effective treatment. One of the most common pathogenic variants is a de novo deep intronic nucleotide substitution in COL6A1 (c.930+189C>T), that creates a donor splice site and drives the insertion of an in-frame pseudoexon. In patient-derived cultured cells, application of antisense oligonucleotides effectively skips the pseudoexon and mitigates the dominant-negative effect that the pseudoexon-containing collagen a1(VI) chain exerts on matrix assembly. To pursue preclinical studies, we generated a unique humanized knock-in (KI) mouse at the Col6a1 gene locus. Here, we present the molecular and phenotypic characterization of this KI model.

Using quantitative PCR and sequencing, we found that the pseudoexon is correctly spliced in. However, it is included in only about 20% of the transcripts from the humanized allele, in gastrocs and diaphragms from 8-week-old males (n=10), which corresponds to half of what is observed in patient muscle biopsies. Nonetheless, we measured reduced muscle mass and weaker grip strength, significant as early as one month-old of age, in both male and female KI animals (p<0.05, n=66).

Comparable to human fibroblasts, cultured skeletal muscle interstitial fibroblasts isolated from KI mice responded efficiently to the PMOs that are designed for human cells (>90% skipping with 5μM treatment). We synthesized our lead antisense sequence on various chemical scaffolds (phosphorothioate 2’methoxyethyl, and vivo-PMO) to compare their in vivo activity. Data collection for systemic injections is ongoing.

In summary, this humanized KI mouse model recapitulates a recurrent splicing defect in COL6A1. Muscle mass and grip strength can be used as outcome measures. The Col6a1/COL6A1 KI mouse represents an invaluable tool to test human-ready antisense oligonucleotides. This work was funded by the MDA and by MDUK.