Transcriptomic analysis of post-traumatic osteoarthritis mouse models identifies miR-199a-5p as a regulator of abnormal mechanical loading in articular cartilage

Objective Animal models of post-traumatic osteoarthritis (PTOA) recapitulate the pathological changes observed in human PTOA. Here we aimed to compare the cartilage transcriptome responses of a non-surgical, mechanically induced rupture of the anterior cruciate ligament (ACL) model and the surgical destabilisation of the medial meniscus (DMM) model. Methods Skeletally mature male C57Bl6 mice were subjected to either the non-surgical, mechanical ACL rupture or surgical DMM models and transcriptome profiling performed on micro-dissected cartilage at day 7 and 42 post-procedure, respectively; in general, naïve animals served as controls. MicroRNA profiling was also performed on the ACL rupture model. Expression levels of a miRNA of interest, miR-199-5p, were inhibited in primary human articular chondrocytes (HAC) with RNA-seq and 3'UTR assays used to identify and valid potential target genes. Results The number of differentially expressed genes between the two models were comparable and highly correlative (Spearman R =0.8, P<2.2E-16). Gene ontology enrichment analysis identified similarly enriched pathways, containing anabolic terms including 'extracellular matrix organisation' enriched for the upregulated genes. Within the ACL rupture miRNA transcriptome, miR-199-5p family members were amongst the most abundantly, and differentially expressed, which was replicated in the DMM cartilage by qRT-PCR. Inhibition of miR-199-5p in HAC led to a comparable transcriptome response to that observed in both human OA damaged vs intact cartilage and murine DMM cartilage datasets. Several genes, including GIT1, NCEH1, SOS2 and ECE1 were all experimentally verified as targets. Conclusion For the first time, we have characterised both the mRNA and miRNA articular cartilage signature in the ACL rupture model and demonstrated highly correlative responses with the DMM PTOA model. These data support the use of the ACL rupture model as a non-invasive alternative to DMM. Overall design: Wild-type C57bl6 mice were used (n=4 per group). Mice were either naive to surgery (D0) or were subjected to destabilisation of the medial meniscus (DMM) surgery (D42 group) and left for disease to develop for 42 days (6 weeks). For the naive group (D0) the mice were 11 weeks of age, for the surgery group (D42)the mice were 17 weeks of age (e.g. 11 weeks plus 42 days). For all mice, Medial knee cartilage caps were dissected, washed with PBS and frozen in liquid nitrogen. Tissue was ground by being placed in an autoclaved chamber with a ball (Retsch) and 250µl QIAzol lysis reagent (QIAGEN). The chambers were transferred to Retsch MM200 mixer mill and tissue ground at vibration frequency of 25 Hz for 90 seconds. To this was added an additional 250µl QIAzol lysis reagent and the mixture transferred to an RNase free tube and incubated at room temperature for 5 minutes. 100µl of chloroform was added and the sample vortexed for 15 s, incubated at room temperature for 10 minutes then centrifuged at 12,500 x g at 4oC for 5 minutes. The upper RNA containing aqueous phase was transferred into a new Eppendorf and RNA and miRNA were purified using a mirVana™ miR Isolation Kit (Ambion, Fisher Scientific, Loughborough, UK) followed by DNAse treatment (DNA-free™ DNA Removal Kit, Invitrogen) following the manufacturer's protocol.