A Col9a3 exon 3 skipping mouse as a novel model for multiple epiphyseal dysplasia

Abstract

A Col9a3 EXON 3 SKIPPING AS NOVEL MODEL FOR MULTIPLE EPIPHYSEAL DYSPLASIA. Introduction: Multiple epiphyseal dysplasia (MED) is generally an autosomal dominant chondrodysplasia characterised by early-onset degenerative joint disease. Its genetic background is complex and heterogeneous and among the mutated genes are those encoding for the pro-α chains of the collagen type IX, COL9A1, COL9A2 and COL9A3, where the majority of mutations lead to the skipping of a syngenic exon. Aim: Reproduce in a mouse model the skip of exon 3 in the COL9A3 gene to understand its pathogenic role in relation to MED. Material and Methods: By CRISPR/Cas9 technology we generated two mouse lines, one carrying a deletion of Col9a3 exon 3 (Col9a3Δex3/Δex3), reproducing the splicing events reported in a MED patients group and a Col9a3-null mouse (Col9a3-/-). On the mutant mice, we performed skeletal X-ray phenotyping and growth plate analysis, including immunohistochemistry and BrdU labelling to monitor proliferation. Ultrastructure of growth plate was visualised using Transmission electron microscopy (TEM). Sequential protein extraction and Atomic force microscopy (AFM) were used to evaluate growth plate cartilage stability, whereas articular cartilage integrity was assessed by destabilisation of the medial meniscus (DMM) surgery and during ageing. Bone density and ultrastructure was assessed by Microcomputed tomography (μCT). Mutant transcriptomes were obtained by RNAseq. Results: The phenotyping of CRISPR/Cas9 generated offspring through DNA and cartilage RNA analysis had led to the establishment of two transgenic mouse lines, one splicing as predicted (Col9a3Δex3/Δex3) and a second lacking the Col9a3 transcript and collagen type IX protein (Col9a3-/-). Both lines are viable, however only Col9a3-/- mice displayed detectable phenotypic abnormalities: mild short stature and hip dysplasia, abnormal tibial epiphysis morphology and delayed ossification of femoral head in 18-week old mice. A reduced level of growth plate chondrocyte proliferation was detected in both mutant mice compared to WT, along with softening of the proliferative zone shown by AFM indentation measurements on new-born and 6 weeks old animals. Interestingly, both mutant mice exhibited similar transcriptome profiles. Only in Col9a3-/- mice immunoblotting of sequentially extracted matrilin-3 and COMP proteins revealed different cartilage extractability from controls. Articular cartilage integrity seemed not to be affected by either mutation. Discussion and conclusion: Analysis confirmed the production of a shorter transcript from cartilage of Col9a3Δex3/Δex3mice. However, these mice still produced collagen type IX protein and had no overt phenotype apart from reduced chondrocytes proliferation and softer proliferative cartilage extracellular matrix. The Col9a3-/- mice had a mild skeletal phenotype and express no Col9a3 transcript or protein, resulting in overall more unstable cartilage. Both mutant mice will represent an important tool to gain insights on collagen type IX role into the matrix. In particular, the Col9a3Δex3/ Δex3 line, by recapitulating human Col9-MED, can add to our understanding of MED disease mechanism.